下载文档原格式
河南省创新发展联盟2024-2025学年高三上学期9月月考英语试题一、阅读理解Join a Zion National Park ranger (护林人) to learn about what makes Zion National Park unique. Programs are free and created for classrooms and individuals. We connect to your school or home through a free web-based program. You will be provided with a link to the video conference ahead of time via an email invite. Registration is open! Click on the program below for more information. Program 1—Chat with a RangerIn Chat with a Ranger, students learn about Zion National Park, the park service, and the life of a ranger. Students prepare and send questions ahead of time. This program can be adapted to fit different curriculum objectives, and is appropriate for any age group. Program 2—Pollination InvestigationIn this distance learning program, students will discover what pollination is and how important it is to all ecosystems. Looking at the relationship between plants and pollinators, participants will see how they have influenced each other and will be challenged to create their own perfect pollinator. Program 3—Whooo’s in the Canyon?Who left these clues behind here in the high canyons of Zion National Park? A feather, small bones, and hoot hooting in the trees can be heard as your classroom goes on a virtual hike of Zion to discover the Mexican spotted owl. Learn it about how the owl uses its special adaptations to survive in this desert environment. Program 4—The Forests, Wetlands, and Deserts of Zion This distance learning program focuses on the plants and animals that live in Zion's varying ecosystems. Students will learn about their adaptations and relationships to each other in this interactive lesson with a creative and critical thinking activity.1．Which program requires participants to make preparations in advance?A．Chat with a Ranger.B．Pollination Investigation.C．Whooo's in the Canyon?D．The Forests, Wetlands, and Deserts of Zion. 2．What can participants learn from program 3?A．Survival strategies taken by owls in the park.B．Ways to prepare a hike tour in the park.C．Threats brought by the desert environment.D．A variety of ecosystems in ZionNational Park.3．What do the listed programs have in common?A．They involve interactive activities.B．They include a virtual tour of different trails.C．They are accessible through web-based program.D．They require participants to visit the park in person.On a hot June day in 2015, I retired after 34 years of teaching high school. Then, I drove to meet my new piano teacher, Mark.I had worked for more than three decades as a busy English teacher with an endless stream of papers to mark and precious little time to experiment or learn new skills. I was determined to make up for all I had been missing. I wanted to finally master the piano and learn how to make music.I told Mark I had a specific concrete goal: to play Clair de lune by Claude Debussy, a piece I remember hearing from early childhood.Determined that there would be a day when I would totally master this piece, I set myself a deadline: I would perform before a gathering of friends on my 60th birthday. For months I did nothing but furiously (猛烈地) practise. When the day came, around 30 friends and relatives crowded into my dining room to hear me play, and aside from a few minor slips, I managed to pull it off without embarrassing myself. People clapped warmly. I made it. I had risen to a challenge, but I still didn’t feel that I was really “making music”.After that, my progress was painfully slow. I had come to hate hearing myself play music badly. I got no pleasure from the act of missing notes.I began focusing on what few things I could do: gardening and cycling. I came to understand that I didn’t have to be that man I’d always thought I ought to be. I could just do what feels good. So, after nearly five years of lessons, I quit.I still love music; I regularly go out to concerts. But now my piano does nothing more than sit silently in my dining room, displaying family photos and collecting dust. And I’m perfectly happy with that.4．Why did the author learn the piano after retiring from teaching?A．To impress his friends and relatives.B．To avoid the boredom of retirement.C．To start a new career as a concert pianist.D．To pursue a long-time passion for music. 5．What can be inferred from paragraph 4?A．The author attended a concert of piano music.B．The author performed successfully despite a few errors.C．The author felt embarrassed about his piano performance.D．The author quit his piano immediately after his 60th birthday.6．What does the author do with his piano now?A．He uses it for music lessons.B．He uses it for performance.C．He uses it for something unrelated to music.D．He plays it for personal enjoyment occasionally.7．Which of the following can best describe the author?A．Inner- directed and hardworking.B．Conventional and careless.C．Ambitious and kind-hearted.D．Lazy and pessimistic.When it comes to diatoms (硅藻类) that live in the ocean, new research suggests that photosynthesis (光合作用) is not the only strategy for accumulating carbon. Instead, these single-celled are also building biomass by feeding directly on organic carbon in the ocean.These new findings could lead researchers to reduce their estimate of how much carbon dioxide diatoms pull out of the air via photosynthesis, which in turn, could take a much closer look at the understanding of the global carbon cycle, which is especially relevant given the changing climate. The new findings were published in Science Advances on July 17, 2024.The team showed that the diatom Cylindrotheca closterium, which is found in oceans around the world, regularly performs a mix of both photosynthesis and direct eating of carbon from organic sources such as plankton (浮游生物) . In more than 70% of the water samples the researchers analyzed from oceans around the world, the team found signs of simultaneous photosynthesis and direct organic carbon consumption from Cylindrotheca closterium. The team also showed that this diatom species can grow much faster when consuming organic carbon in addition to photosynthesis. Furthermore, the new research hinted at the possibility that specificspecies of bacteria are feeding organic carbon directly to a large percentage of these diatoms living all across the global ocean. This work is based on a genome-scale metabolic modeling approach that the team used to reveal the metabolism of the diatom Cylindrotheca closterium.The team’s new metabolic modeling data support recent lab experiments suggesting that some diatoms may rely on strategies other than photosynthesis to intake the carbon they need to survive, thrive and build biomass.The UC San Diego led team is in the process of expanding the scope of the project to determine how widespread this non-photosynthetic activity is among other diatom species. 8．What’s new according to the research?A．The way of the diatom’s carbon accumulation.B．The impact of climate on diverse sea plants.C．The procedure of exploring carbon.D．The system of building biomass.9．What do the new findings make researchers more focus on?A．The causes of climate change.B．The grasp of the carbon cycle.C．The bad effect of photosynthesis on diatoms.D．A rough estimate of the amount of carbon dioxide.10．What do we know from paragraph 3?A．A large number of diatoms may feed on bacteria.B．The diatom lives on plankton.C．Water samples are key factors for the research.D．Diatom species grow faster with sufficient sunlight11．Which is the most suitable title for the text?A．Photosynthesis in Diatoms B．Plankton’s Role in OceansC．New Carbon Strategies in Diatoms D．Advances in Modeling DataAccording to a report in 2023, the World Health Organization (WHO) recommended that non-sugar sweeteners not be used as a means of achieving weight control or reducing the risk of diseases. The guideline came as a surprise. After all, the very purpose of non-sugar sweeteners-which contain little to no calories—is to help consumers control their weight and reduce their risk of disease by replacing sugar.In its report, the WHO cited evidence that long-term use of non-sugar sweeteners is associated with an increased risk of diabetes (糖尿病) and death. How is it that non-sugar sweeteners are linked to the negative health effects they’re supposed to fend off?The WHO made its recommendation after reviewing hundreds of published studies. The problem is that the overwhelming majority of these studies are observational. In such studies, subjects tend to self-report their food intake, which might not guarantee inaccuracy. More importantly, observational studies cannot determine cause and effect. Are non-sugar sweeteners causing diabetes, or are people at risk of diabetes simply more likely to consume them? Lastly, there are numerous variables that researchers can’t possibly control for in these studies that could influence the results.Randomized controlled trials (RCTs) tell a different story about non-sugar sweeteners. These studies control for variables by randomly assigning people to either a treatment or control group, and they can determine cause and effect. They show that sweeteners modestly benefit weight loss and help control blood sugar, without the negative effects seen in observational research. The downside of RCTs is that they are shorter in duration, often lasting just a few months. So negative effects could appear after longer use and we wouldn’t be able to tell from these RCTs.But we also can’t tell from observational studies, which only measure correlation and not causality (因果关系) . Changing the current situation might be hard, though. RCTs are expensive and require recruiting participants, setting up diet plans, and regularly measuring subjects’ health outcomes.For change to happen, it might need to start at the top, where science is funded Government agencies, which appropriate billions for research, should start prioritizing RCTs.12．What do the underlined phrase “fend off” probably mean in paragraph 2?A．Put out.B．Defend against.C．Keep up.D．Count on. 13．What does paragraph 3 mainly talk about?A．The WHO’s suggestions on observational studies.B．The strategies to decide cause and effect in conducting studies.C．The significance of controlling variables in observational studies.D．The limitations of the observational studies in the WHO report.14．What is a feature of RCTs according to the text?A．They cost little B．They tend to last long.C．They can control variables and determine causality.D．They require participants to self-report related data15．How should the government help RCTs?A．By making appropriate plans B．By providing financial supportC．By raising people’s awareness of health D．By founding more related governmentagenciesTo make science’s stories more concrete and engaging, it’s important to use some effective strategies. Here are four of them. Put people in the storyScience’s stories often lack human characters. 16 . Characters can be also people affected by a scientific topic, or interested in learning more about it. Besides, they can be storytellers who are sharing their personal experiences.17People often think of science as objective and fair. But science is actually a human practice that continuously involves choices, missteps and biases (偏见) . If you explain science as a course, you can walk people through the sequence of how science is done and why researchers reach certain conclusions. 18 . And they can also stress the reason why people should trust the course of science to provide the most accurate conclusions possible given the available information. Include what people care aboutScientific topics are important, but they may not always be the public’s most pressing concerns. In April 2024, a polling company found that “the quality of the environment” was one of thelowest-ranked priorities among people in the US. The stories about the environment could weave in connections to higher-priority topics. 19 . Tell science's storiesScientists, of course, can be science communicators, but everyone can tell science’s stories. When we share information online about health, or talk to friends and family about the weather, we contribute to information that circulates about science topics. 20 . Think about all of a story’s characteristics - character, action, sequence, scope, storyteller and content - and how you might incorporate them into the topic.A．Explain science as a processB．Shoot attractive short science videosC．Scientists themselves can actually become ideal onesD．This practice is to stress why the content is importantE．You can tell growth stories of remarkable teenage scientistsF．Science communicators can emphasize how science is conductedG．You may as well borrow features from stories to strengthen your message二、完形填空In 2018, Molly Baker unfortunately lost her husband in a severe skiing accident. She was 21 . In the first several weeks after his passing, her friends and family 22 a great deal of support. But after a while, the cards and meals started to 23 . “People had to get back to their normal 24 . And so things kind of dropped off,” Baker recalled.That was when one of Baker's friends, Carla Vail, thought up a way to 25 the help for an entire year. She called it the “Calendar Girls”. V ail gathered the names of 31 of Baker's friends who wanted to help, and 26 each friend a particular day. Vail also gave Baker the names on the 27 , so Baker could know what to 28 each day.“And what that looked like for them was that on that day, they would reach out to me in some 29 ways—maybe via text, or a card,” Baker said.Looking back, Baker feels that Vail's 30 was essential to helping her cope with her husband's death, because she was 31 at that time.“A lot of people are really uncomfortable around 32 ,” Baker said. “So what they do is, instead of doing something, that they 33 do nothing. It was nice to have that ‘Calendar Girls’ setup.”Today, Baker tries to do something similar for her friends going through 34 . In hard times, she knows how 35 it is to have something to look forward to every day. 21．A．cautious B．unconscious C．desperate D．impassive 22．A．extended B．demanded C．announced D．assumed 23．A．pass down B．show up C．break up D．slow down24．A．exercise B．routine C．diet D．growth 25．A．resist B．continue C．explain D．test 26．A．ordered B．sent C．owed D．assigned 27．A．furniture B．file C．calendar D．Internet 28．A．expect B．absorb C．propose D．define 29．A．rare B．strange C．specific D．generous 30．A．curiosity B．thoughtfulness C．ambition D．toughness 31．A．innocent B．optimistic C．tolerant D．lonely 32．A．panic B．evidence C．failure D．grief 33．A．simply B．hardly C．skillfully D．secretly 34．A．distraction B．addiction C．loss D．annoyance 35．A．amusing B．valuable C．astonishing D．universal三、语法填空阅读下面短文，在空白处填入1个适当的单词或括号内单词的正确形式。
题目Argument150The following is a letter to the editor of an environmental magazine."The decline in the numbers of amphibians worldwide clearly indicates the global pollution of water and air. Two studies of amphibians in YosemiteNational Park in California confirm my conclusion. In 1915 there were seven species of amphibians in the park, and there were abundant numbers of each species. However, in 1992 there were only four species of amphibians observed in the park, and the numbers of each species were drastically reduced. The decline in Yosemite has been blamed on the introduction of trout into the park's waters, which began in 1920 (trout are known to eat amphibian eggs). But the introduction of trout cannot be the real reason for the Yosemite decline because it does not explain the worldwide decline."直译如下：全球范围内两栖动物数量的减少证明全球范围内空气和水的污染。
2023年12月英语六级阅读原文原文标题：The Importance of Environmental Protection随着工业化和城市化的进程不断加快，环境问题已经成为全球性的焦点。
关于这一话题，许多人有不同的看法。
有一些人认为环境污染是制约人类社会发展的主要障碍之一，应当尽快加强环境保护。
而也有一些人对此持怀疑和否定态度，认为环境问题并不严重，环境污染对人类社会的发展不构成实质性的威胁，因此不必大惊小怪。
无论如何，我们都不能忽视环境问题的存在和严重性。
环境保护的重要性首先体现在生态系统的稳定和人类生存环境的改善。
生态系统是地球上的重要基础设施，不同的生物之间通过各种复杂的生态关系相互依存，形成生态系统的稳定性。
然而，由于人类活动过度开发和环境污染，使得原有的生态系统遭到破坏，生物多样性下降，一些濒临灭绝的物种濒临灭绝。
这对于人类生存环境产生了严重的影响。
加强环境保护，保持生态系统的稳定性，保护生物多样性是十分必要的。
环境保护对于人们的身体健康和身心健康是至关重要的。
环境污染直接危害人们的身体健康，长期暴露在污染环境中会导致各种慢性疾病的发生，甚至致癌。
而且，环境污染还会对人们的心理健康造成影响，长期生活在污染环境中会使人产生消极的情绪，降低人们的生活质量。
为了保护人们的身体健康和身心健康，必须加强环境保护，减少环境污染。
环境保护与可持续发展的理念相一致，是现代社会发展的必由之路。
可持续发展是指经济、社会和环境的协调发展，即在满足当前需求的前提下，能够保证子孙后代也能满足其需要。
如果环境得不到有效的保护，将严重威胁人类社会的可持续发展，甚至会导致资源过度消耗，生态平衡被打破，给子孙后代留下巨大的环境债务。
加强环境保护，推动可持续发展已成为全球热点问题。
环境保护对于人类社会的发展至关重要。
只有加强环境保护，才能保障生态系统的稳定和人类生存环境的改善，保护人们的身体健康和身心健康，推动可持续发展。
湖北省六校2025届高三英语11月联考试题（含解析）留意事项:1.答题前，先将自己的姓名、准考证号填写在试题卷和答题卡上，并将准考证号条形码粘贴在答题卡上的指定位置。
2.选择题的作答:每小题选出答案后，用2B铅笔把答题卡上对应题目的答案标号涂黑，写在试题卷、草稿纸和答题卡上的非答题区域均无效。
3.非选择题的作答:用签字笔干脆答在答题卡上对应的答题区域内。
写在试题卷、草稿纸和答题卡上的非答题区域均无效。
4.考试结束后，请将试卷和答题卡按时上交。
第一节听下面 5 段对话。
每段对话后有一个小题，从题中所给的 A、B、C 三个选项中选出最佳选项。
听完每段对话后，你都有 10 秒钟的时间来回答有关小题和阅读下一小题。
每段对话仅读一遍。
1. What’s the relationship between the man and Ms. Jones?A. Husband and wife.B. Teacher and student.C. Neighbors.2. Where did the man stay last night?A. At a classroom.B. At a party.C. In a lab.3. When did the lecture probably start?A. At 10: 00 a. m.B. At 9: 00 a. m.C. At 8: 30 a. m.4. What will the woman probably do next?A. Go to the man’s place.B. Call the Hillsboro Hotel.C. Reserve an exhibition hall.5. What’s the man’s favorite sport at the 2024 Beijing Winter Olympics?A. Alpine skiing.B. Curling.C. Figure skating.其次节听下面 5 段对话或独白。
参考译文如何应对全球流行病布莱恩•沃什[1]对于墨西哥城这样拥有两千万人口的熙熙攘攘的大都市而言，恢复生机的第一个标志不是公开的宗教仪式，也不是政治集会，而是交通拥堵，这很正常。
为应对甲型H1N1流感的爆发，墨西哥城实施了为期一周的停工停市，终于在五月五日这天重归繁华与喧闹。
猪流感的传播速度已经减缓，这使得墨西哥的政府官员燃起希望，觉得最糟糕的时刻已经过去了。
[2]国际卫生组织的官员也可以稍感放松了。
自从四月底首次出现新型流感病毒报告以来，他们一直处于高度紧张状态。
世界卫生组织和疾病预防控制中心的科学家发现在墨西哥以外鲜有严重或致死的甲型H1N1流感病例，另外也没有足够证据显示疾病会在多数国家持续传播。
[3]那么，世界范围内的关闭学校、边境检测和政府要员召开新闻发布会号召民众勤洗手是大惊小怪吗？很可惜，不是。
正如卫生组织的官员反复强调的，我们现在还处于甲型H1N1流感爆发的初期，而流感病毒的不可预测性是众所周知的。
目前，这种新型流感病毒似乎和普通季节性流感的危险性相差无几，但它可能会在明年冬天以更为致命的方式卷土重来——正如1918年那场造成灾难性后果的大流感一样。
[4]现实情况是，虽然美国和其他国家的卫生部门官员在应对甲型H1N1流感病毒方面进行的广泛合作值得赞扬，但H1N1病毒的出现还只是敲响一个警钟，并不是对我们意志和能力的真正考验。
哥伦比亚大学的美国国家灾难防御中心主任爱尔文•莱德纳尔博士指出：“我们应该把这种新型流感病毒的出现看作是警钟，而不是闹铃。
”[5]在甲型H1N1流感病毒面前，我们这个广泛联系的全球社会对新出现的疾病所显示出的无助暴露无遗。
乘飞机旅行和国际贸易往来使得新型病菌能在不到两周的时间内传播到二十几个国家。
当然，全球化有不利因素也有其优势。
正是全球化使我们能够建立起一个真正意义上的全球疾病监测系统。
全球性流行病的威胁提醒美国人必须对自己老朽过时的医疗卫生系统进行修正。
在传染病爆发的时候，每个人的处境都岌岌可危。
A hidden Markov model for the detection of pure andmixed strategy play in games∗Jason Shachat†J.Todd Swarthout‡Lijia Wei§July7,2012AbstractWe propose a statistical model to assess whether individuals strategically use mixed strategies in repeated games.We formulate a hidden Markov model in which the latentstate space contains both pure and mixed strategies,and allows switching between thesestates.We apply the model to data from an experiment in which human subjectsrepeatedly play a normal form game against a computer that always follows its part ofthe unique mixed strategy Nash equilibrium profile.Estimated results show significantmixed strategy play and non-stationary dynamics.We also explore the ability of themodel to forecast action choice.JEL classification:C92;C72;C10Keywords:Mixed Strategy;Nash Equilibrium;Experiment;Hidden Markov Model ∗This paper supersedes the previous working paper,“Man versus Nash:An experiment on the self enforcing nature of mixed strategy equilibrium.”†Wang Yanan Institute for Studies in Economics and MOE Key Laboratory in Econometrics,Xiamen University.jason.shachat@‡Department of Economics and Experimental Economics Center,Georgia State University. swarthout@§Wang Yanan Institute for Studies in Economics and MOE Key Laboratory in Econometrics,Xiamen University.ljwie.wise@1IntroductionGame theory and the Nash equilibrium solution concept are a key framework in the social sciences for modeling interactive behavior.The formulation of a normal form game consists of a set of players,a set of possible actions for each player,and a payofffunction for each player that gives a real-valued payofffor any possible joint action profile–a list of actions consisting of one for each player.A Nash equilibrium is a joint action profile such that each player’s assigned action results in at least as high a payoffto the player as any other possible action,assuming all other players choose their respective actions in the Nash equilibrium profile.If players are restricted to deterministically choose an action,then there are many games that don’t have a Nash equilibrium,such as the childhood game of Rock,Scissors, Paper.Confronted with this problem,Von Neumann(1928)generalized a player’s decision from choosing an action to choosing a probability distribution over his possible actions.1This choice of a probability distribution is called a“mixed”strategy,and a degenerate mixed strategy which chooses a particular action with probability one is called a“pure”strategy. The introduction of mixed strategies allows for existence of equilibrium across a broad class of games:from minimax solutions for zero-sum games(Von Neumann,1928;Von Neumann and Morgenstern,1944)to noncooperative equilibria for n-person games(Nash,1951).While the role of mixed strategies in defining logically consistent solution concepts is not in doubt, the positive aspect of individuals actually playing mixed strategies is an open question of considerable interest.Researchers’efforts to answer this question have naturally focused on settings where the use of mixed strategies is most compelling:the repeated play of games which have a unique mixed strategy Nash equilibrium.The value of“being unpredictable”is readily seen in examples such as serves in tennis,“bluffing”in poker,and whether or not a tax authority audits a tax payer.A common approach in this literature is to test whether the players’action choices are consistent with the mixed strategy equilibrium.Some studies using controlled experiments with human subjects have the advantage of knowing the payofffunctions,and test whether choice frequencies agree with the equilibrium strategies and whether players’sequences of actions are serially independent(O’Neill,1987;Binmore,Swierzbinski,and Proulx,2001; Morgan and Sefton,2002;Selten and Chmura,2008).Other studies consider high-level sports competitions,such as soccer(Chiappori,Levitt,and Groseclose,2002;Palacios-Huerta,2003; Bareli,Azar,Ritov,Keidarlevin,and Schein,2007)and tennis(Walker and Wooders,2001), with the advantage of studying highly experienced players competing for high stakes and 1Along with generalizing the set of feasible actions to the set of mixed strategies,a player’s payofffunction is extended by setting its value to the expected payoffgiven a profile of mixed strategies,commonly referred to as the expected utility property.the disadvantage of unknown payofffunctions.2These studies focus on testing the serial independence of action choice and the equilibrium implication of equal payoffs across action choices.Some of the most prominent and recurring results for both types of studies are that aggregated action frequencies across players agree with the equilibrium mixed strategies but individual action frequencies do not,and for many individuals action choices are serially correlated violating the independence prediction.To reconcile these issues of serial correlation and heterogeneity,several studies(Ochs,1995; Bloomfield,1994;Shachat,2002;Noussair and Willinger,2011)conduct laboratory experi-ments using the same type of games but directly elicit mixed strategies by obligating players to select a probability distribution over actions.3Elicited strategies in these experiments ex-hibit various distinct patterns.Some subjects choose pure strategies almost exclusively,some choose strictly mixed strategies almost exclusively,and others use both types of strategies–usually in long sequences.Also,certain mixed strategies are often quite focal,such as choosing equal probability weight on a subset of actions rather than the Nash equilibrium proportions. Naive interpretation of these results suggests a clear distinction between play that is purposely unpredictable and play that is a pure best response to changing forecasts of an opponent’s action(Nyarko and Schotter,2002).A more cautious interpretation is that subjects may es-chew the randomizing device provided by the experimenter and instead internally randomize, or perhaps subjects choose strictly mixed strategies due to the experimenter effect of the novel elicitation method.Clearly a less invasive method to detect mixed strategy play would be valuable.In this study we propose a hidden Markov model(HMM)to detect whether observed action choices are the result of pure or mixed strategies play in repeated two-personfinite action games.4There are three key ideas in our formulation:(1)we treat the strategy a player follows as a latent state and the action played as the observable output from the latent strategy;(2) the set of possible latent states is a discrete subset of all possible mixed strategies containing pure strategies,Nash equilibrium or minimax strategies,and focal mixed strategies;and(3) a player switches the latent strategy he follows according to afirst order Markov process. We then demonstrate the ability of the model by applying it to a new experimental data set we collect.In our experiment,each human subject repeatedly plays a2×2game against a computer player that follows its mixed strategy equilibrium.Some subjects play a zero-2The action sets are typically comprised of simple actions,e.g.,{serve left,serve right}and{defend left, defend right}.The payoffs are assumed to be the probability of winning the task and these probabilities will differ based upon both the comparative skills between players and the relative strengths a player has for each action.3For example,Shachat(2002)adopts a game with four actions,each identified by a different color,for each player.Each player mustfill a box with100cards in any combination of the four colored card types,and then one card is selected at random to determine the action played.4See Rabiner(1989)for a classic introduction to hidden Markov models.sum game and others an unprofitable game.5The estimated HMMs reveal several interesting results,including:(1)significant amounts of both pure and mixed strategy play;(2)the focal equiprobable mixed strategy is played more often than the Nash equilibrium strategy;(3)low transition probabilities between mixed and pure latent strategies;(4)dynamic adjustments in the types of strategies players follow over time;and(5)appreciable rates of both mixed and pure strategy play in the limiting distributions of the HMMs(interpreted as the long run equilibrium of play).We then extend the HMM from a statistical framework for evaluating hypotheses to one for forecasting action choice and assess its predictive accuracy.2A HMM of switching strategiesConsider an experiment in which we observe M pairs of subjects,each playing T periods of the same2×2normal form game.Often games like this are described by a two-by-two table,and for familiarity purposes we denote one subject’s player role as Row and the other as Column. We label each player role’s two possible actions Left(L)and Right(R),and express a subject’s mixed strategy as the probability of playing L.Of particular interest is when the game has a single Nash equilibrium and it is in strictly mixed strategies,although our framework is not restricted to study only such cases.Three factors confounding the analysis data generated by this type of process are the latency of players’mixed strategies,the heterogeneity of strategy adoption across subjects,and variation of adopted latent strategies over the course of repeated play.In this section,we present a model that accommodates and allows estimation of these confounds.Consider the following HMM for afixed player role.The state space S is an n-element subset of the subject i’s possible mixed strategies.Denote s i,t∈S for the strategy used by subject i in period t,S i is the set of all possible T element sequences of mixed strategies for i with typical element s i,and let s be the collection of s i for all M subjects in a given player role.Let y i,t denote subject i’s realized action in period t,y i is the corresponding T element sequence of i’s observable actions,and y is the collection of y i for all M subjects.View{y,s} as the output of the HMM.The probability structure of the HMM has three elements.First,the n-element vector B for which the element B j is the probability a subject chooses action Left,i.e.the mixed strategy,if he is in state j.We will provide two analyses which differ in how we specify B.In one approach we consider B as known a priori,and S and B are redundant ually, in this approach,B contains the two pure strategies,other strategies suggested by theory such 5An unprofitable game is one in which the minimax and Nash equilibrium solutions are distinct but yield the same expected payofffor each player.as Nash equilibrium or minimax,and other focal strategies.In the second approach we treat the elements of B as unknown parameters–the state dependent mixed strategies.The second element of the structure,π,is the initial multinomial probability distribution over S.The third element,P,is the n×n transition probability matrix.The element P jk is the probability a subject adopts strategy k in period t conditional upon having adopted strategy j in period t−1.The likelihood function of(B,π,P)isL(B,π,P|y,s)=Pr(y,s|B,π,P).Rewriting this likelihood in terms of the marginal distributions of y and s gives usL(B,π,P|y,s)=Pr(y|s,B,π,P)·Pr(s|B,π,P).Next,we assume that the marginal distribution of y conditional on s is independent ofπand P.In other words,once the state is realized then the probability of a Left action relies solely on the mixed strategy of the current state.Also,by the specification of the HMM,s is independent of the state dependent mixed strategies B.This allows us to restate the previous likelihood function asL(B,π,P|y,s)=Pr(y|s,B)·Pr(s|π,P).Since the sequence of states for each subject is unobservable,we evaluate the likelihood by integrating over the set of all possible sequencesL(B,π,P|y,s)=M i=1s∈Sπ(s i,1)B I y i,1s i,1(1−B si,1)1−I y i,1Tt=2P si,t−1,s i,tB I y i,ts i,t(1−B si,t)1−I y i,t ,where I y i,t is an indicator function which equals one for the action Left and zero for the action Right.As T grows,the number of calculations needed to evaluate this likelihood quickly becomes computationally impractical.We describe the Bayesian approach we take to estimate the HMM,although one could proceed down a frequentist path of maximizing the expected likelihood function using some variation of the EM(expected maximum likelihood) algorithm.In the Bayesian analysis,wefirst factor the joint posterior distribution of the unknown HMM parameters and unobserved states s into the product of marginal conditional posterior distributions.Then we evaluate these marginal conditional posteriors through an iterative sampling procedure called the Markov Chain Monte Carlo(MCMC)method.MCMC is asimple but powerful procedure in which the empirical distributions of the sampled parameters converge to the true posterior distributions.After convergence,iterative sampling is continued to construct empirical density functions.These are used to make inferences regarding the parameters of the hidden Markov models.Consider the posterior density function on the realized unobserved states and HMM pa-rameters h(s,B,P,π|y).First,express this joint density as the product of the marginal density of HMM parameters conditional on the observed action choices and unobserved states with the marginal density of the states conditional upon action choicesh(s,B,P,π|y)=h(B,P,π|s,y)h(s|y).We have already assumed that the transition matrix P and initial probabilities over statesπare independent of the action choices and state contingent mixed strategies B,which allows us to stateh(s,B,P,π|y)=h(B|s,y)h(P,π|s,y)h(s|y).This product of three conditional posteriors permits a simple Markov Chain procedure of sequentially sampling from these distributions.We start with some initial arbitrary values for the HMM parameters,(B(l),P(l),π(l))where l=0.We create s(0)by simulation using P(0)and π(0)without conditioning on y.From these initial parameter values and the observed action sequences y,we use a Gibbs sampling algorithm to generate an initial sample of state sequences s(1).Then we make a random draw P(1)from the posterior distribution of P conditional on s(1)and y,and proceed similarly to make a random draw ofπ(1).We complete the iteration by making a random draw B(1)from the posterior of B conditional on s(1)and y.The key to the MCMC method is that as l→∞,the joint and marginal distributions of B(l),P(l), andπ(l)converge weakly to the joint and marginal posterior distributions of these parameters (Geman and Geman,1987).We now describe the details of each step in an iteration of the MCMC procedure.Step1:Sampling the state sequences s(l)We begin by describing a Gibbs sampling technique for generating draws from the distribu-tion of s(l)conditional upon y and(B(l−1),P(l−1),π(l−1)).The elements of s i can be drawn sequentially for each t conditioning on the observed action choice y i,t,the realized state in other periods,π,and P.Let s i,=t be the vector obtained by removing s i,t from the sequence s i.Given s i,=t,we express the conditional posterior distribution of s i,t asPr(s(l)i,t |y i,t,B(l−1),P(l−1),s(l)i,=t,s(l−1)i,=t)∝Pr(y i,t|s(l)i,t,B(l−1))·Pr(s(l)i,t|P(l−1),s(l)i,=t,s(l−1)i,=t)Pr(s(l)i,t |P(l−1),s(l)i,=t,s(l−1)i,=t)=Pr(s i,t=k|P(l),s(l)i,t−1,s(l−1)i,t+1).Consequently,the conditional posterior probability of s i,t=k and t>1isPr(s(l)i,t =k|·)=Pr(y i,t|s i,t=k,B(l−1)k)·Pr(s i,t=k|P(l−1),s(l)i,t−1,s(l−1)i,t+1) nj=1Pr(y i,t|s i,t=j,B(l−1)j)·Pr(s i,t=j|P(l−1),s(l)i,t−1,s(l−1)i,t+1),and for t=1Pr(s(l)i,1=k|·)=Pr(y i,1|s i,1=k,B(l−1)k)·Pr(s i,1=k|π(l−1),s(l−1)i,2) nj=1Pr(y i,1|s i,1=j,B(l−1)j)·Pr(s i,1=j|π(l−1),s(l−1)i,2).The state s(l)i,tis determined by making a random draw from the uniform distribution on theunit interval,and comparing this draw to the calculated conditional probability of s(l)i,t. Step2:Sampling the transition matrix P(l)andπ(l)The posterior distributions of P jk andπdepend only upon s(l)and the priors.We specify the prior ofπas a Dirichlet distribution h(π;α1,...,αn)whereαj=1,for1≤j≤n.Similarly, we specify the prior of the j th row of P as a Dirichlet distribution h(p j1,...,p jn|ηj1,...,ηjn). In an experiment,we record the data from the true start of the HMM process,so we assume that the joint posterior is simply the product of these two marginal posteriors.The respective posteriors ofπ(l)and P(l)areh(π|s)∝Pr(s|π)h(π;α1,...,αn),andh(P j1,...,P jn|s)∝Pr(s|P j1,...,P jn)h(P j1,...,P jn;ηj1,...,ηjn).Ifν0j is the number incidences of s(l)i,1=j in s(l),andνjk is the count of transitions from state j to k in s(l),then the conditional probabilities in the two posterior calculations are multinomial distributionsh(π|s)∝πν011...πν0n−1n−1·1−n−1k=1πkν0nh(π;α1,...,αn)h(P j1,...,P jn|s)∝Pνj1j1...Pνjn−1jn−1·1−n−1k=1P jkνjnh(P j1,...,P jn;η1,...,ηn).Since the Dirichlet distribution is the conjugate prior for the multinomial distribution,these posterior distributions are also Dirichlet distributions for which each shape parameter is the sum of its prior value and the respective counth(π|s)=h(π;α1+ν01,...,αn+ν0n)andh(P j1,...,P jn|s)=h(P j1,...,P jn;η1+νj1,...,ηn+νjn).Hence,we selectπ(l)and P(l)be taking random draws from these distributions.Step3:Sampling the state dependent mixed strategies BFor our initial approach to modeling the state dependent mixed strategies,we assume B corresponds to a known subset of S.In our Bayesian analysis this is equivalent to assuming a point prior on these strategies,and therefore there is no updating.So in our Gibbs sampling procedure we skip this step,and proceed to next iteration of the Gibbs sampler.Of course this is a rather strong prior to assume,and we should evaluate whether it is appropriate. Accordingly,we conduct an auxiliary analysis in which we assume a uniform prior of the set of all mixed strategies.In the auxiliary analysis we proceed as follows.The priors of state dependent mixed strate-gies B1,...,B n are assumed independent of each other and of the Markov process governing the states.Given these assumptions,we can think of each B j as a Bernoulli probability, and each Left(Right)action as a success(failure)when occurring in state j.The likelihood function is calculated as a binomial trial.Since it is the conjugate prior of the binomial,we assume the prior is a Beta distribution,denotedβ(B j;ζj;γj).We want a uniform prior as well,and that corresponds to setting the shape parametersζj andγj to one.The posterior distribution is simplyh(B j|y,s(l))=β(B j;ζj+κL,j,γj+κR,j),whereκL,j andκR,j are the number of times the actions Left and Right,respectively,are chosen when in state j according to s(l).The state conditional mixed strategies B(l)j,j=1,...,n, are randomly drawn from these Beta posterior distributions,completing an iteration of theGibbs sampler.The Gibbs sampler is run for a large number of iterations until the empirical distribution of all the parameters has converged(Geweke,1991).Then the sampling procedure is allowed to continue to run for another number of iterations to build up an empirical distribution that corresponds to the posterior distribution of the HMM parameters.It is from this empirical distribution that we conduct statistical inferences.3The experimentWe apply our HMM framework to a new experimental data set that provides a likely setting for mixed strategies,and particulary Nash equilibrium strategies,to be adopted.Additionally, our procedures allow us to estimate for one player role without the need to also simultaneously model the opposing role,because each human subject repeatedly plays against a computer player that follows its mixed strategy equilibrium.Each subject is informed that his opponent is a computer but is given no information regarding the computer’s strategy.We adopt two different games in our experimental design,with each subject playing only one of the two games.One game is zero-sum and the other game is unprofitable.3.1The gamesOurfirst game is a zero-sum asymmetric matching pennies game introduced by Rosenthal, Shachat,and Walker(2003).The normal form representation of this game is presented on the left side of Figure1.The game is called Pursue-Evade because the Row player“captures”points from the Column player when the actions of the two players match,and the Column player“evades”a loss when the players’actions differ.In the game each player can move either Left or Right,and the game has a unique Nash equilibrium in which each player chooses Left with probability two-thirds.In equilibrium,Row’s expected payoffis two-thirds, and correspondingly Column’s expected payoffis negative two-thirds.Our second game is an unprofitable game introduced by Shachat and Swarthout(2004) called Gamble-Safe.Each player has a Gamble action(Left for each player)which yields a payoffof either two or zero,and a Safe action(Right for each player)which guarantees a payoffof one.The normal form representation of this game is presented on the right side of Figure1.The Gamble-Safe game has a unique Nash equilibrium in which each player chooses the Left action with probability one-half,and each player earns an expected equilibrium payoffof one.Right is the minimax strategy for both players with a guaranteed payoffof one.Aumann(1985)argues that the Nash equilibrium prediction is not plausible in such an unprofitable game because its adoption assumes unnecessary risk to achieve the correspondingPursue-Evade Game Gamble-Safe Game Column Player Left RightR o w P l a y e r Left 1 , -1 0 , 0 Right 0 , 0 2 , -2Column PlayerLeft Right R o w P l a y e r Left 2 , 0 0 , 1Right 1 , 2 1 , 1 Figure 1:The experimental gamesNash equilibrium payoff.For example,imagine Row has Nash equilibrium beliefs and best responds by playing the Nash strategy.Row’s expected payoffis one.However,suppose Column instead adopts his minimax strategy Right.This reduces Row’s expected payoffto one-half.Row could avoid this risk by simply playing the minimax strategy.This aspect makes the Gamble-Safe game a more challenging test for the hypothesis of mixed strategy play than the zero-sum Pursue-Evade game.3.2Subject recruitment and experiment protocolWe conducted six experiment sessions in the Finance and Economics Experimental Laboratory (FEEL)at Xiamen University during December 2011.A total of 110undergraduate and masters students participated in the experiment,with each session containing between 12and 22subjects.54subjects were assigned to the Persue-Evade game treatment,and 56subjects were assigned to the Gamble-Safe game treatment.Subjects were evenly divided into Row and Column player roles within each treatment.FEEL uses the ORSEE online recruitment system for subject recruitment (Greiner,2004),and at the time of the experiment approximately 1400students were in the subject pool.A subset of students from the subject pool were invited to attend each specific session,and these students were informed that they would receive a 10Yuan show-up payment and have the opportunity to earn more money during the experiment.Further,the invitation stated that the session would last no more than two hours.Upon arrival at the laboratory,each subject was seated at a computer workstation such that no subject could observe another subject’s screen.Subjects first read instructions de-tailing how to enter decisions and how earnings were determined.6Then,200repetitions of the game were played.For the Pursue-Evade game,Column subjects were initially endowed with a balance of 260tokens each,and Row subjects none.Each token was worth one-third6The instructions are available at /swarthout/HMM/of a Yuan.Each subject’s total earnings consisted of the10Yuan show-up payment plus the monetary value of his token balance after the200th repetition.While a mathematical possibility,no Column subjects in the Pursue-Evade game went bankrupt.The experiment was conducted with a Java software application created at the Georgia State University Experimental Economics Center(E x CEN)that allows humans to play normal form games against computerized algorithms.At the beginning of each repetition,each subject saw a graphical representation of the game on the screen.Each Column subject’s game display was transformed so that he appeared to be a Row player.Thus,each subject selected an action by clicking on a row,and then confirmed his choice.After the repetition was complete,each subject saw the outcome highlighted on the game display,as well as a text message stating both players’actions and his own earnings for that repetition.Finally,each subject’s current token balance and a history of past play were displayed at all times.The history consisted of an ordered list with each row displaying the repetition number,the actions selected by both players,and the subject’s payofffrom the specific repetition.3.3Data summaryWe begin the summary of the experimental data by providing views of the joint distribution of the proportion of Left play for each subject-computer pair,while conditioning on whether the data are from thefirst100or last100repetitions.Figures2and3present these views for the Pursue-Evade and Gamble-Safe treatments,respectively.In each of thesefigures,the x-axis is the proportion of Left play for the Column player and the y-axis is the proportion of Left play for the Row player.Each arrow in thefigures represents the play of a single human-computer pair,with the arrow tail representing the joint frequency of Left play in the first100repetitions and the arrow head representing the joint frequency of Left play in the final100repetitions.These arrows show the adjustments subjects make from thefirst half to the second half of play.We see that many arrows suggest substantial change in the human player frequency,but the changes do not trend in any one direction or uniformly towards the Nash equilibrium.Human play also displays greater dispersion and displacement from the Nash equilibrium than the computer opponents,suggesting nonconformity with the Nash equilibrium predictions.Table1presents the means and standard deviations of subjects’frequencies of Left play by treatment and role.Recall that we have2700observations for the each role in the Pursue-Evade treatment and2800observations for each role in the Gamble-Safe treatment.Although the Row player mean is close to the Nash equilibrium proportion in both game treatments,the Nash equilibrium proportion is rejected for all four cases at any reasonable level of significance. In each of the four cases,subjects’proportions of Left play display too much variance to haveHuman Row vs.NE ColumnNE Row vs.Human ColumnComputer Column Proportion LeftH u m a n R o w P r o p o r t i o n L e f t0.00.20.40.60.81.00.00.20.40.60.81.0C o m p u t e r R o w P r o p o r t i o n L e f t0.00.20.40.60.81.00.00.20.40.60.81.0Computer Column Proportion LeftH u m a n R o w P r o p o r t i o n L e f t0.00.20.40.60.81.00.00.20.40.60.81.0C o m p u t e r R o w P r o p o r t i o n L e f t0.00.20.40.60.81.00.00.20.40.60.81.0。
参考译文伴生物种1. 伴生物种是指不被计算在上岸渔获量中的，但是受到捕捞影响的物种。
跨界鱼类种群，高度洄游鱼类和公海鱼类种群因受到如下因素的影响而影响其他物种：（1）丢弃，（2）未被捕捞上来的生物与渔具发生身体接触，（3）间接过程。
2. 渔业通过很多种机制来影响伴生物种，丢弃是目前人类获取知识最多的一种，尽管人类所知有限。
关于丢弃的全球最新信息是一份粮农组织的报告。
该报告估计全球海洋渔业的丢弃率约为百分之八，丢弃率会根据不同的国家，齿轮类型，目标物种和统计区发生改变。
3. 虾类拖网作业的平均丢弃率最高，为百分之六是二点三。
不同渔业的丢弃率差别很大，在零到百分之九十六之间变化。
尽管有一些跨界的或其他公海中的虾类种群的捕捞，大多数虾类拖网作业仍然限于对专属经济区中虾类种群的捕捞。
专属经济区中虾类的捕捞目标很有可能是生活在较深水域或冷水水域的物种。
冷/深水水域捕虾业的总丢弃率是百分之三十九，但在使用副渔获减少装置（BRDs）后（比如在格林兰岛），丢弃率相对较低，在百分之五左右。
混获的有各种长须鲸和无脊椎动物物种，也包括其他渔业中目标物种的幼鱼。
对于虾类拖网作业中丢弃的长须鲸物种（尤其是比目鱼）的关注促使一些渔业强制使用副渔获减少装置（BRDs）。
4. 延绳钓捕捞高洄游鱼种（主要是金枪鱼和类金枪鱼属物种）具有仅次于虾类拖网作业的丢弃率（平均丢弃率为百分之二十八，并且在零百分之四十范围内浮动）。
延绳钓中最常见的丢弃物种是蓝鲨。
其他鲨鱼，受到鲨鱼和海洋哺乳动物损害的目标物种，扁舵鲣，鲔，印度洋国王鲭鱼，和土魠鱼也在被捕获后丢弃。
5. 跨界鱼类种群和公海鱼类种群主要采用底层拖网捕捞。
目标为底栖鱼类的拖网渔船丢弃率百分之九点六（所有渔业）。
没有根据来判断跨界鱼类种群和公海鱼类种群的丢弃率与专属经济区鱼类种群的丢弃率孰高孰低。
专属经济区鱼类种群的捕捞量在总捕捞量中所占的比例如此之高以至于研究者估计目标为底栖鱼类的底层拖网捕捞所导致的1，700，000吨丢弃物大多来源于专属经济区渔业。
A Clear Statement of the Novelty of the ResearchIntroductionThe introduction gives an overview of the importance of novelty in research and sets the context for the rest of the article.The Significance of Novelty in Research1.The role of novelty in advancing knowledge2.Importance of identifying gaps in existing research3.Novelty as a driver of scientific progressDefinition of Novelty in Research1.The concept of novelty2.Types of novelty in research–Methodological novelty–Theoretical novelty–Empirical noveltyResearch Methodology1.Research design and approach2.Data collection methods3.Data analysis techniques–Descriptive analysis–Inferential analysisIdentification of Novelty in the Research1.Novelty in research objectives2.Novelty in research questions3.Novelty in research methodology–Integration of multiple methodologies–Use of innovative techniques4.Novelty in research findings–Unexpected results–New insights or perspectivesAssessment of Novelty in the Research1.Peer review process–Importance of peer review in evaluating novelty–Criteria used by reviewers to assess novelty2.Citation analysis–Identification of previous works cited–Evaluation of the extent of novelty in citationsImpact of Novelty in Research1.Contribution to existing knowledge2.Influence on future research directions3.Relevance for practical applications or policy-makingConclusionThe conclusion highlights the importance of novelty in research and emphasizes the need for continuous innovation and exploration in scientific endeavors.Through this comprehensive analysis, it becomes evident that the novelty of research plays a fundamental role in advancing scientific knowledge and driving progress. Identifying and incorporating novelty within research can lead to groundbreaking discoveries and significant contributions to various fields. By understanding the different aspects of novelty and employing rigorous methodologies, researchers can enhance the impact and relevance of their work, ultimately promoting the growth and development of their respective disciplines.。
初三英语完形填空深度理解单选题60题1. In the classic novel, the character was in a dilemma. He had to ______ between staying with his family and pursuing his dream.A. chooseB. selectC. electD. pick答案：A。
解析：choose是普通用词，侧重根据个人意愿和判断从众多的对象中进行选择，这里表达在家庭和梦想之间做出选择，比较通用。
select较正式，强调经过认真考虑后的挑选，通常是从多个类似事物中进行挑选，在这里语境没有那么正式。
elect主要用于选举，是选举某人担任某职位，不符合语境。
pick通常用于口语，有挑选、采、摘等意思，不如choose正式且常用于这种两难抉择的语境。
2. The scientific article mentioned that the new species ______ a unique feature that distinguishes it from others.A. hasB. ownsC. possessesD. holds答案：C。
解析：possess表示拥有，常指拥有抽象的东西，如品质、特征等，这里说新物种拥有独特的特征，用possess最合适。
has 是最普通的表示“有”，比较口语化。
own强调合法地拥有某物，多指所属关系，如拥有财产等，这里不是指所属关系。
hold表示握住、持有（具体东西）或者举行（会议等），不符合这里表示拥有特征的语境。
3. In the adventure story, the hero ______ his courage when facing the dangerous situation.A. showedB. displayedC. exhibitedD. demonstrated答案：A。
