DOCSLIB.ORG

New Technology Assessment in Entrepreneurial Financing-Can

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
New Technology Assessment in Entrepreneurial Financing-Can New Technology Assessment in Entrepreneurial Financing - Can Crowdfunding Predict Venture Capital Investments? Jermain Kaminski,∗ Christian Hopp,y Tereza Tykvovaz Draft, August 26, 2018 Abstract 1 Introduction Recent years have seen an upsurge of novel sources of new venture financing through crowdfunding (CF). We draw on 54,943 successfully crowdfunded projects "[T]his year, we've been slower to invest partially be- and 3,313 venture capital (VC) investments through- cause in our analysis, there are years where there are out the period 04/2012-06/2015 to investigate, on lots of new ideas and big swings that are going for the aggregate level, how crowdfunding is related to new industries. I feel like last year and maybe the a more traditional source of entrepreneurial finance, year before were better years for big new ideas. This venture capital. Granger causality tests support year, we haven't seen as many."1 the view that VC investments follow crowdfunding investments. Cointegration tests also suggest a Aileen Lee, Cowboy Ventures long-run relationship between crowdfunding and VC investments, while impulse response functions (IRF) indicate a positive effect running from CF to VC Entrepreneurship is \always a voyage of explo- within two to six months. Crowdfunding seems to ration into the unknown, an attempt to discover new help VC investors in assessing future trends rather ways of doing things better than they have been done than crowding them out of the market. before" Hayek (1948, p. 101). The need for experi- mentation is deeply rooted in entrepreneurship. An Keywords environment which facilitates experimentation and Crowdfunding, venture capital, Granger causality, tolerates failures is the mainstay of a prospering eco- crowding out. nomic ecosystem that generates innovations (Kerr, Nanda, and Rhodes-Kropf, 2014). Classification Within such an environment, many new business G30, G24, O3 ideas compete, but only a few single ventures will arXiv:1608.07182v1 [cs.CY] 25 Aug 2016 actually survive and bring their products to market. As future outcomes are distant und unpredictable, entrepreneurs and their financiers are plagued with uncertainty about whether or not a particular prod- uct will sell or a business model will turn out to be successful. Different types of financiers deal with ∗RWTH Aachen, Technology Entrepreneurship Group; these uncertainties in different ways. In crowdfund- MIT Media Lab, Macro Connections Group. ing, many small investors place bets on the potential yRWTH Aachen, Technology Entrepreneurship Group. zHohenheim University, Chair of Corporate Finance; Cen- 1http://techcrunch.com/2015/09/28/early-stage- tre for European Economic Research (ZEW). investors-apply-the-brakes; accessed June 1st, 2016. 1 of a business idea. To receive venture capital (VC) market, but also provides VC investors with a crowd- funding business ideas undergo a profound selection driven proof of concept for certain technologies and process through a specialized investor. The aim of markets. Just like bees may indicate the direction this paper is to investigate how crowdfunding relates to a distant flower patch yielding nectar and pollen to VC. More specifically, we seek to answer the ques- through a `waggle dance' in their hive (Von Frisch, tion whether crowdfunding and VC are rather com- 1967), crowdfunding investors may fulfil a comple- plements or substitutes. mentary role to VC investors, by providing them with Crowdfunding is a relatively new gateway for en- information towards new products and product cat- trepreneurs to access capital for creative and novel egories. Crowdfunding platforms may thus support ideas (Belleflamme, Lambert, and Schwienbacher, products and ideas, which would otherwise not be 2014; Cumming and Johan, 2016; Mollick, 2014; pursued. The \wisdom of the crowd" may not only Schwienbacher and Larralde, 2012). It allows indi- lead to more projects that search for VC financing, viduals and creative users to start experiments with but also improve the quality of projects that are in new products and services (Von Hippel, 1986) within the market. This is important as deal flow has been a wide range of capital supply. The amount can vary shown as a large concern for VC investors (Hochberg, from a few dollars for a pottery project to $20 mil- Ljungqvist, and Lu, 2010), in particular in ho peri- lion for a new smart watch. Average individual con- ods when \too much money chases too few deals" tributions are rather low2 and entrepreneurial ven- (Gompers and Lerner, 2000). Alternatively, we con- tures typically use crowdfunding in the very early, sider the opposite view, namely that crowdfunding exploratory phase. The money is raised via online and VC investors compete for deals in some segments platforms, such as Kickstarter, which provide en- of the market, where the availability of crowdfunding trepreneurs with a direct interaction with potential increases the supply of funds that compete for a lim- end users. Through their feedback, end users provide ited number of projects. If this is the case, crowd- evaluations of the business idea and may prove use- funding and VC investments will be substitutes and ful in reshaping and improving the product. The en- crowdfunding would crowd out VC investments. trepreneurs not only attract potential customers and Our research questions ask whether there is a long- early-stage financiers on the crowdfunding platform, run relationship between crowdfunding and VC in- but also obtain higher visibility, for example through vestments on the aggregate and the industry levels, press coverage. Crowdfunding may help to aggregate and whether crowdfunding and VC investments are market expectations about the likelihood of future linked, such that an increase in crowdfunding causes events, trends and technologies. Mollick and Kup- an increase (or decrease) in VC investments. To puswamy (2014) suggest that crowdfunding appears answer these questions, we use a dataset covering to support traditional entrepreneurship. In fact, if 54,943 crowdfunding investments that have success- the crowd provides expert-like assessment of future fully reached their funding goal and 3,313 VC in- technology trends, than a positive long-run relation- vestments in the US. Our analysis starts with the ship between crowdfunding and VC investments may adoption of JOBS Act in April 2012, which fueled exist, in which VC investments follow crowdfunding crowdfunding investments, and ends in June 2015. investments at a time lag. We test whether one time series Granger causes the Crowdfunding processes may validate the feasibil- other, and we employ a cointegration test to investi- ity and viability of new technologies. Crowdfunding gate whether there is a long-run relationship between not only reflects an investment into future products, crowdfunding and VC investments. The results sug- i.e., needs that may not be satisfied in the current gest that crowdfunding Granger causes VC invest- ments. We also find a long-term relationship be- 2 Mollick (2014) reports the average goal of $5,064 for all tween crowdfunding and VC investments; the time se- 2,317 funded projects on Kickstarter in the period from its in- ception in 2009 to July, 2012; the average contribution amounts ries are cointegrated. Successful crowdfunding cam- to $20,563 in our sample. paigns lead to a subsequent increase in VC invest- 2 ments. This holds in aggregate and especially for It is not only the quantity, but the early test for hardware and consumer electronics. viability may also improve the quality of projects that compete in the marketplace. Crowdfunded en- We perform a separate analysis for `small' VC in- trepreneurs might generate spillovers and spur the vestments below $500,000 (which are comparable to birth and development of further companies in the crowdfunding investments in their size) to investi- same and other technology fields as successful crowd- gate whether crowdfunding may potentially crowd funding campaigns can create interest in new projects out VC investments within this segment. We do and co-creations. not find evidence in support of a crowding-out ef- fect. Our work helps to clarify the role that crowd- funding plays in financing entrepreneurial ventures. 2 Related Literature and Research It contributes to several strands of the literature. Question We add to the emerging literature on crowdfunding (e.g. Agrawal, Catalini, and Goldfarb, 2015; Ahlers et 2.1 Entrepreneurship as Experimentation al., 2015; Lehner, Grabmann, and Ennsgraber, 2015; Cumming and Johan, 2016; Mollick, 2013; Mollick In March 2014, when Facebook bought Oculus Rift and Kuppuswamy, 2014; Mollick and Nanda, 2015). { a company which raised money via Kickstarter {, To the best of our knowledge, no empirical analy- Marc Zuckerberg stated: \One day, we believe this sis exists that focuses on the possible connection be- kind of immersive, augmented reality will become a tween crowdfunding and VC investments, though re- part of daily life for billions of people. Virtual reality cent work addresses the link between business an- was once the dream of science fiction. But the inter- gels and VC investors (Berkovitch, Grinstein, and Is- net was also once a dream, and so were computers rael, 2015; Chemmanur and Chen, 2014; Hellmann, and smartphones."3 Schure, and Vo, 2015; Hellmann and Thiele, 2015). At the origin of such radically new products or ser- vices, there are creative and innovative entrepreneurs, Our study extends previous analyses that sug- who, instead of being
Load more

Recommended publications
  • A Hierarchical Bayesian Model for Improving Wisdom of the Crowd
    A hierarchical Bayesian model for improving wisdom of the crowd aggregation of quantities with large between-informant variability Saiwing Yeung ([email protected]) Institute of Education, Beijing Institute of Technology, China Abstract ties can be described by distributions without a natural maxi- The wisdom of the crowd technique has been shown to be very mum and are severely right skewed. For example, Gibrat’s effective in producing judgments more accurate than those of law suggested that the distribution of populations of cities individuals. However, its performance in situations in which follows a log-normal distribution (Eeckhout, 2004). Other the intended estimates would involve responses of greatly dif- fering magnitudes is less well understood. We first carried out distributions with similar characteristics include power-law, an experiment to elicit people’s estimates in one such domain, Pareto, and exponential distributions. They naturally occur in populations of U.S. metropolitan areas. Results indicated that many different contexts, including income and wealth, num- there were indeed vast between-subjects differences in magni- tudes of responses. We then proposed a hierarchical Bayesian ber of friends, waiting time, time till failure, etc. (Barabási, model that incorporates different respondents’ biases in terms 2005). How to best aggregate these quantities in a WoC con- of the overall magnitudes of their answers and the amount of text is not very well understood. In the present research we individual uncertainties. We implemented three variations of this model with different ways of instantiating the individual demonstrate a hierarchical Bayesian approach to the problem. differences in overall magnitude. Estimates produced by the Hierarchical Bayesian models formally express the rela- variation that accounts for the stochasticities in response mag- tionships between psychological constructs, stimuli, and ob- nitude outperformed those based on standard wisdom of the crowd aggregation methods and other variations.
    [Show full text]
  • A Statistical Model for Aggregating Judgments by Incorporating Peer Predictions
    A statistical model for aggregating judgments by incorporating peer predictions John McCoy3 and Drazen Prelec1;2;3 1Sloan School of Management Departments of 2Economics, and 3Brain & Cognitive Sciences Massachusetts Institute of Technology, Cambridge MA 02139 [email protected], [email protected] March 16, 2017 Abstract We propose a probabilistic model to aggregate the answers of respondents answering multiple- choice questions. The model does not assume that everyone has access to the same information, and so does not assume that the consensus answer is correct. Instead, it infers the most prob- able world state, even if only a minority vote for it. Each respondent is modeled as receiving a signal contingent on the actual world state, and as using this signal to both determine their own answer and predict the answers given by others. By incorporating respondent’s predictions of others’ answers, the model infers latent parameters corresponding to the prior over world states and the probability of different signals being received in all possible world states, includ- ing counterfactual ones. Unlike other probabilistic models for aggregation, our model applies to both single and multiple questions, in which case it estimates each respondent’s expertise. The model shows good performance, compared to a number of other probabilistic models, on data from seven studies covering different types of expertise. Introduction It is a truism that the knowledge of groups of people, particularly experts, outperforms that of arXiv:1703.04778v1 [stat.ML] 14 Mar 2017 individuals [43] and there is increasing call to use the dispersed judgments of the crowd in policy making [42].
    [Show full text]
  • Social Learning, Strategic Incentives and Collective Wisdom: an Analysis of the Blue Chip Forecasting Group
    Social Learning, Strategic Incentives and Collective Wisdom: An Analysis of the Blue Chip Forecasting Group J. Peter Ferderer Department of Economics Macalester College St. Paul, MN 55105 [email protected] Adam Freedman Chicago, IL 60601 [email protected] July 22, 2015 ABSTRACT: Using GDP growth forecasts from the Blue Chip survey between 1977 and 2011, we measure absolute consensus errors and forecast dispersion for each of the 24 monthly forecast horizons and provide three main findings. First, the size of consensus errors and forecast dispersion are negatively correlated over longer-term forecast horizons (from 24 to 13 months). This finding is consistent with the hypothesis that the Lawrence Klein forecasting award, which is based on performance of 12-month-head forecasts, increased the group’s collective wisdom by raising the incentive to anti-herd. Second, absolute consensus errors and forecast dispersion display significant negative temporal variation for the longest horizons (24 to 20 months). Third, after the early 1990s (i) there was a dramatic decline in forecast dispersion associated with a significant increase in the size of longer-term consensus errors, and (ii) forecasts bracket realized GDP growth much less frequently. The final two results suggest that increased herding or reduced model diversity caused collective wisdom to diminish in the second part of the sample. JEL Classifications: E37 Forecasting and Simulation: Models and Applications D70 Analysis of Collective Decision-Making, General D83 Search; Learning; Information and Knowledge; Communication; Belief Keywords: social learning, herding, strategic incentives, reputation, consensus forecasts, collective wisdom, forecast errors, forecast dispersion, Great Moderation 0 The key lesson I would draw from our experience is the danger of relying on a single tool, methodology or paradigm.
    [Show full text]
  • Debiasing the Crowd: How to Select Social
    1 Debiasing the crowd: how to select social 2 information to improve judgment accuracy? 3 Short title: How to select social information to improve judgment accuracy? 1;∗ 2 1 4 Bertrand Jayles , Cl´ement Sire , Ralf H.J.M Kurvers 1 5 Center for Adaptive Rationality, Max Planck Institute for Human Development, Lentzeallee 94, 6 14195 Berlin, Germany 2 7 Laboratoire de Physique Th´eorique,Centre National de la Recherche Scientifique (CNRS), 8 Universit´ede Toulouse { Paul Sabatier (UPS), Toulouse, France 9 Abstract 10 Cognitive biases are widespread in humans and animals alike, and can sometimes 11 be reinforced by social interactions. One prime bias in judgment and decision making 12 is the human tendency to underestimate large quantities. Former research on social 13 influence in estimation tasks has generally focused on the impact of single estimates on 14 individual and collective accuracy, showing that randomly sharing estimates does not 15 reduce the underestimation bias. Here, we test a method of social information sharing 16 that exploits the known relationship between the true value and the level of under- 17 estimation, and study if it can counteract the underestimation bias. We performed 18 estimation experiments in which participants had to estimate a series of quantities 19 twice, before and after receiving estimates from one or several group members. Our 20 purpose was threefold: to study (i) whether restructuring the sharing of social infor- 21 mation can reduce the underestimation bias, (ii) how the number of estimates received 22 affects sensitivity to social influence and estimation accuracy, and (iii) the mechanisms 23 underlying the integration of multiple estimates.
    [Show full text]
  • Studying the ``Wisdom of Crowds'' at Scale
    Studying the “Wisdom of Crowds” at Scale Camelia Simoiu,1 Chiraag Sumanth,1 Alok Mysore,2 Sharad Goel1 1 Stanford University, 2University of California San Diego Abstract a county fair. He famously observed that the median of the guesses—1,207 pounds—was, remarkably, within 1% of the In a variety of problem domains, it has been observed that the true weight (Galton 1907). aggregate opinions of groups are often more accurate than those of the constituent individuals, a phenomenon that has Over the past century, there have been dozens of studies been dubbed the “wisdom of the crowd”. However, due to that document this “wisdom of crowds” effect (Surowiecki the varying contexts, sample sizes, methodologies, and scope 2005). Simple aggregation—as in the case of Galton’s ox of previous studies, it has been difficult to gauge the extent competition—has been successfully applied to aid predic- to which conclusions generalize. To investigate this ques- tion, inference, and decision making in a diverse range tion, we carried out a large online experiment to systemati- of contexts. For example, crowd judgments have been cally evaluate crowd performance on 1,000 questions across 50 topical domains. We further tested the effect of different used to successfully answer general knowledge ques- types of social influence on crowd performance. For exam- tions (Surowiecki 2005), identify phishing websites and ple, in one condition, participants could see the cumulative web spam (Moore and Clayton 2008; Liu et al. 2012), crowd answer before providing their own. In total, we col- forecast current political and economic events (Budescu lected more than 500,000 responses from nearly 2,000 par- and Chen 2014; Griffiths and Tenenbaum 2006; Hill and ticipants.
    [Show full text]
  • Crowd-Sourced Data Coding for the Social Sciences: Massive Non-Expert Coding of Political Texts*
    Crowd-sourced data coding for the social sciences: massive non-expert coding of political texts* Kenneth Benoit Drew Conway London School of Economics New York University and Trinity College, Dublin Michael Laver Slava Mikhaylov New York University University College London Abstract A large part of empirical social science relies heavily on data that are not observed in the field, but are generated by researchers sitting at their desks, raising obvious issues of both reliability and validity. This paper addresses these issues for a widely used type of coded data, derived from the content analysis of political text. Comparing estimates derived from multiple “expert” and crowd-sourced codings of the same texts, as well as other independent estimates of the same latent quantities, we investigate whether we can analyze political text in a reliable and valid way using the cheap and scalable method of crowd sourcing. Our results show that, contrary to naive preconceptions and reflecting concerns often swept under the carpet, a set of expert coders is also a crowd. We find that deploying a crowd of non-expert coders on the same texts, with careful specification and design to address issues of coder quality, offers the prospect of cheap, scalable and replicable human text coding. Even as computational text analysis becomes more effective, human coding will always be needed, both to validate and interpret computational results and to calibrate supervised methods. While our specific findings here concern text coding, they have implications for all expert coded data in the social sciences. KEYWORDS: text coding; crowd sourcing; expert-coded data; reliability; validity * Prepared for the third annual New Directions in Analyzing Text as Data conference at Harvard University, 5-6 October 2012.
    [Show full text]
  • Counteracting Estimation Bias and Social Influence To
    bioRxiv preprint doi: https://doi.org/10.1101/288191; this version posted March 24, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. Counteracting estimation bias and social influence to improve the wisdom of crowds Albert B. Kaoa,∗, Andrew M. Berdahlb, Andrew T. Hartnettc, Matthew J. Lutzd, Joseph B. Bak-Colemane, Christos C. Ioannouf, Xingli Giamg, Iain D. Couzind,h aDepartment of Organismic and Evolutionary Biology, Harvard University, Cambridge, MA, USA bSanta Fe Institute, Santa Fe, NM, USA cArgo AI, Pittsburgh, PA, USA dDepartment of Collective Behaviour, Max Planck Institute for Ornithology, Konstanz, Germany eDepartment of Ecology and Evolutionary Biology, Princeton University, Princeton, NJ, USA fSchool of Biological Sciences, University of Bristol, Bristol, UK gDepartment of Ecology and Evolutionary Biology, University of Tennessee, Knoxville, TN, USA hChair of Biodiversity and Collective Behaviour, Department of Biology, University of Konstanz, Konstanz, Germany Abstract Aggregating multiple non-expert opinions into a collective estimate can improve accuracy across many contexts. However, two sources of error can diminish collective wisdom: individual esti- mation biases and information sharing between individuals. Here we measure individual biases and social influence rules in multiple experiments involving hundreds of individuals performing a classic numerosity estimation task. We first investigate how existing aggregation methods, such as calculating the arithmetic mean or the median, are influenced by these sources of error. We show that the mean tends to overestimate, and the median underestimate, the true value for a wide range of numerosities. Quantifying estimation bias, and mapping individual bias to collective bias, allows us to develop and validate three new aggregation measures that effectively counter sources of collective estimation error.
    [Show full text]
  • An Experiment About the Impact of Social Influence on the Wisdom of the Crowds' Effect
    School of Technologies and Architecture, ISCTE Faculty of Sciences, University of Lisbon An experiment about the impact of social influence on the wisdom of the crowds' effect Sofia Silva ([email protected]) Dissertation submitted in partial requirement for the conferral of Master of Sciences degree in Complexity Sciences Supervisor: Professor Luís Correia Faculty of Sciences, University of Lisbon June 2016 Acknowledgments I would like to express my profound gratitude to my dearest friend István Mandak, without whom this dissertation would have not been possible. He contributed with his knowledge and ideas to implement the code for the experiment and committed his time to improve the way data was collected. Preventing duplicates in the framework of MTurk has turned out to be exceptionally challenging and his long term support and patience was invaluable to achieve a reliable setup. I equally wish to thank his wife, Adrienn and his two kids, Maja and Bohus, who inadvertently accompanied this project and with whom I ended sharing dinner, games and a great amount of jellybeans. I would like to sincerely thank my supervisor Luís Correia who patiently guided me throughout this bumpy and rather long journey. I believe that without his continuous support, patience, and readiness to communicate (even if three thousand kilometres away) it would have been extremely difficult to overcome the many challenges of this project. His continuous enthusiasm and good humour were truly motivating. I also want to thank Jorge Louçã for his initial encouragement and support when I first started the program. Furthermore, I would like to thank my dearest friends Sabrina Amendoeira, Diana Ferreira, and Juergen Krasser who have always been encouraging and supportive.
    [Show full text]
  • Leveraging Wisdom of the Crowd for Decision Support
    Leveraging Wisdom of the Crowd for Decision Support Simo Hosio Jorge Goncalves Theodoros Anagnostopoulos Vassilis Kostakos University of Oulu University of Oulu University of Oulu University of Oulu Pentti Kaiteran katu 1 90570 Pentti Kaiteran katu 1 90570 Pentti Kaiteran katu 1 90570 Pentti Kaiteran katu 1 90570 Oulu, Finland Oulu, Finland Oulu, Finland Oulu, Finland [email protected] [email protected] [email protected] [email protected] While Decision Support Systems (DSS) have a long history, their usefulness for non-experts outside specific organisations has not lived up to the promise. A key reason for this is the high cost associated with populating the underlying knowledge bases. In this article, we describe how DSSs can leverage crowds and their wisdom in constructing knowledge bases to overcome this challenge. We also demonstrate how to construct DSSs on-the-fly using the collected data. Our user-driven laboratory studies focus on user perceptions of the concept itself, and motives for contributing and using such a DSS. To the best of our knowledge, our approach and implementation are the first to demonstrate such use of crowdsourcing in building DSSs. Decision Support Systems, evaluation, crowdsourcing, wisdom of the crowd 1. INTRODUCTION Ultimately, our vision is to transform Wisdom of the Crowd into useful DSSs on-the-fly. In doing so, we Decision Support Systems (DSS) (Druzdzel, Flynn hope to overcome several acknowledged 1999) typically combine multiple data sources, limitations in traditional DSSs. expert input, and computational methods, to explore a given problem domain and ultimately help choose from a set of defined options.
    [Show full text]
  • The Consensus Handbook Why the Scientific Consensus on Climate Change Is Important
    The Consensus Handbook Why the scientific consensus on climate change is important John Cook Sander van der Linden Edward Maibach Stephan Lewandowsky Written by: John Cook, Center for Climate Change Communication, George Mason University Sander van der Linden, Department of Psychology, University of Cambridge Edward Maibach, Center for Climate Change Communication, George Mason University Stephan Lewandowsky, School of Experimental Psychology, University of Bristol, and CSIRO Oceans and Atmosphere, Hobart, Tasmania, Australia First published in March, 2018. For more information, visit http://www.climatechangecommunication.org/all/consensus-handbook/ Graphic design: Wendy Cook Page 21 image credit: John Garrett Cite as: Cook, J., van der Linden, S., Maibach, E., & Lewandowsky, S. (2018). The Consensus Handbook. DOI:10.13021/G8MM6P. Available at http://www.climatechangecommunication.org/all/consensus-handbook/ Introduction Based on the evidence, 97% of climate scientists have concluded that human- caused climate change is happening. This scientific consensus has been a hot topic in recent years. It’s been referenced by presidents, prime ministers, senators, congressmen, and in numerous television shows and newspaper articles. However, the story of consensus goes back decades. It’s been an underlying theme in climate discussions since the 1990s. Fossil fuel groups, conservative think-tanks, and political strategists were casting doubt on the consensus for over a decade before social scientists began studying the issue. From the 1990s to this day, most of the discussion has been about whether there is a scientific consensus that humans are causing global warming. As the issue has grown in prominence, a second discussion has arisen. Should we even be talking about scientific consensus? Is it productive? Does it distract from other important issues? This handbook provides a brief history of the consensus on climate change.
    [Show full text]
  • Pre Test Excerpt
    UC Merced Proceedings of the Annual Meeting of the Cognitive Science Society Title Wisdom of the Crowds in Minimum Spanning Tree Problems Permalink https://escholarship.org/uc/item/8z73d04t Journal Proceedings of the Annual Meeting of the Cognitive Science Society, 32(32) Authors Yi, Sheng Kung Steyvers, Mark Lee, Michael et al. Publication Date 2010 Peer reviewed eScholarship.org Powered by the California Digital Library University of California Wisdom of the Crowds in Minimum Spanning Tree Problems Sheng Kung Michael Yi ([email protected]) Mark Steyvers ([email protected]) Michael D. Lee ([email protected]) Department of Cognitive Sciences, University of California, Irvine Irvine, CA 92697, USA Matthew Dry ([email protected]) Discipline of Pharmacology University of Adelaide Adelaide, SA 5005, Australia Abstract found that the median estimate for the weight of the ox was within 1% of the ox‟s actual weight. Similarly, Surowiecki The „wisdom of the crowds‟ effect describes the finding that (2004) reports that, when polled, the modal answer given by combining responses across a number of individuals in a the audience in the US version of the game show “Who group leads to aggregate performance that is as good as or better than the performance of the best individuals in the Wants To Be A Millionaire” for multiple choice questions is group. Here, we look at the wisdom of the crowds effect in correct more than 90% of the time. the Minimum Spanning Tree Problem (MSTP). The MSTP is Recently, the wisdom of the crowds idea has also been an optimization problem where observers must connect a set applied to more complex problems.
    [Show full text]
  • Wisdom of the Crowd for Fault Detection and Prognosis
    Wisdom of the Crowd for Fault Detection and Prognosis Yuantao Fan Supervisors: Thorsteinn Rögnvaldsson Sławomir Nowaczyk DOCTORAL THESIS | Halmstad University Dissertations no. 67 Wisdom of the Crowd for Fault Detection and Prognosis © Yuantao Fan Halmstad University Dissertations no. 67 ISBN 978-91-88749-42-0 (printed) ISBN 978-91-88749-43-7 (pdf) Publisher: Halmstad University Press, 2020 | www.hh.se/hup Printer: Media-Tryck, Lund Abstract Monitoring and maintaining the equipment to ensure its reliability and avail- ability is vital to industrial operations. With the rapid development and growth of interconnected devices, the Internet of Things promotes digitization of in- dustrial assets, to be sensed and controlled across existing networks, enabling access to a vast amount of sensor data that can be used for condition monitor- ing. However, the traditional way of gaining knowledge and wisdom, by the expert, for designing condition monitoring methods is unfeasible for fully uti- lizing and digesting this enormous amount of information. It does not scale well to complex systems with a huge amount of components and subsys- tems. Therefore, a more automated approach that relies on human experts to a lesser degree, being capable of discovering interesting patterns, generating models for estimating the health status of the equipment, supporting mainte- nance scheduling, and can scale up to many equipment and its subsystems, will provide great benefits for the industry. This thesis demonstrates how to utilize the concept of "Wisdom of the Crowd", i.e. a group of similar individuals, for fault detection and prognosis. The approach is built based on an unsupervised deviation detection method, Consensus Self-Organizing Models (COSMO).
    [Show full text]