Archive

Archive for the ‘Stanford University’ Category

Book: Mining of Massive Datasets

January 13, 2014 Comments off

Mining of Massive Datasets
Source: Stanford University

At the highest level of description, this book is about data mining. However, it focuses on data mining of very large amounts of data, that is, data so large it does not fit in main memory. Because of the emphasis on size, many of our examples are about the Web or data derived from the Web. Further, the book takes an algorithmic point of view: data mining is about applying algorithms to data, rather than using data to “train” a machine-learning engine of some sort. The principal topics covered are:

1. Distributed file systems and map-reduce as a tool for creating parallel algorithms that succeed on very large amounts of data.

2. Similarity search, including the key techniques of minhashing and locality- sensitive hashing.

3. Data-stream processing and specialized algorithms for dealing with data that arrives so fast it must be processed immediately or lost.

4. The technology of search engines, including Google’s PageRank, link-spam detection, and the hubs-and-authorities approach.

5. Frequent-itemset mining, including association rules, market-baskets, the A-Priori Algorithm and its improvements.

6. Algorithms for clustering very large, high-dimensional datasets.

7. Two key problems for Web applications: managing advertising and rec- ommendation systems.

8. Algorithms for analyzing and mining the structure of very large graphs, especially social-network graphs.

Hat tip: Research Buzz

About these ads

National Charter School Study 2013

June 25, 2013 Comments off

National Charter School Study 2013 (PDF)
Source: CREDO at Stanford University
From press release (PDF):

A new , independent national study finds improvement in the overall performance of charter schools, driven in part by the presence of more high – performing charters and closure of underperforming charter schools.

The National Charter School Study 2013, released today by the Center for Research on Education Outcomes (CREDO) at Stanford University , is an update and expansi on of CREDO’s 2009 landmark 16-state study , Multiple Choice, the first study to take a comprehensive look at the impact of charter schools on student performance. The 2009 study found a wide variance in quality among charter schools, with students in charter schools not faring as well in the aggregate as those attending traditional public schools.

The National Charter School Study 2013 looks at performance of students in charter schools in 26 states and New York City, which is treated separately as the city differs dramatically from the rest of the state. In those states (and New York City), charter school students now have greater learning gains in reading than their peers in traditional public s chools. Traditional public schools and charter schools have equivalent learning gains in mathematics.

In the aggregate, charter school students in the 26 states in the new study gain ed an additional 8 days of learning each year in readi ng beyond their local peers in traditional public schools. The 2009 study found a loss of 7 days each year in reading among the students in the 16 states. In mathematics, charter school students in 2009 posted 22 fewer days of learning than their traditional public school counterparts; today there exists no significant difference in days of learning.

Executive Summary
Tech Appendix
Supplementary Findings

The Infographics
Virtual Control Records
Interactive Map of Charter States
16-State Update School Groups
16-State Reading State Charter Impact Changes
16-State Math State Charter Impact Changes
27-State School Level Reading Quality
27-State School Level Math Quality
27-State Reading State Charter Impacts
27-State Math State Charter Impacts

Expanding College Opportunities for High-Achieving, Low Income Students

April 12, 2013 Comments off

Expanding College Opportunities for High-Achieving, Low Income Students (PDF)

Source: Stanford Institute for Economic Policy Research

Only a minority of high-achieving, low-income students apply to colleges in the same way that other high-achieving students do: applying to several selective colleges whose curriculum is designed for students with a level of achievement like their own. This is despite the fact that selective colleges typically cost them high-achieving, low-income students less while offering them more generous resources than the non-selective postsecondary institutions they mainly attend. In previous work, we demonstrate that the vast majority of high-achieving, low-income students are unlikely to be reached by traditional methods of informing students about their college opportunities since such methods require the students to be concentrated geographically. In this study, we use a randomized controlled trial to evaluate interventions that provide students with semi-customized information on the application process and colleges’ net costs. The interventions also provide students with no-paperwork application fee waivers. The ECO Comprehensive (ECO-C) Intervention costs about $6 per student, and we find that it causes high-achieving, low-income students to apply and be admitted to more colleges, especially those with high graduation rates and generous instructional resources. The students respond to their enlarged opportunity sets by enrolling in colleges that have stronger academic records, higher graduation rates, and more generous resources. Their freshman grades are as good as the control students’, despite the fact that the control students attend less selective colleges and therefore compete with peers whose incoming preparation is substantially inferior. Benefit-to-cost ratios for the ECO-C Intervention are extremely high, even under the most conservative assumptions.

Monitoring Risks Before They Go Viral: Is it Time for the Board to Embrace Social Media?

April 23, 2012 Comments off
Source:  Stanford Graduate School of Business
Given the pervasiveness of social media, should the board of directors pay closer attention to the information exchanged on these sites? Can this information be used to improve oversight and risk management?

+ Full Document(PDF)

Sudden Death of a CEO: Are Companies Prepared When Lightening Strikes?

April 21, 2012 Comments off
Source:  Stanford University Graduate School of Business
It is very difficult for shareholders to know detailed information about CEO succession planning among the companies they have invested in.  Although CEO deaths are rare, the sudden death of a CEO can provide insight into the quality of succession planning and governance of a company.  Whereas some companies are able to appoint a successor immediately, others take weeks or months to do so.
In this Closer Look, we examine this issue in detail.
We ask:
  • Why haven’t more companies done a “reality check” on whether they have a truly operational succession plan?
  • What can a board learn and what should it do if the market reacts positively to the death of its CEO?
  • Should the board revise its succession plan if its CEO engages in risky hobbies or lifestyle habits?

Repositioning Dynamics and Pricing Strategy

February 17, 2012 Comments off

Repositioning Dynamics and Pricing Strategy
Source: Stanford Graduate School of Business

We measure the revenue and cost implications to supermarkets of changing their price positioning strategy in oligopolistic downstream retail markets. Our estimates have implications for long-run market structure in the supermarket industry, and for measuring the sources of price rigidity in the economy. We exploit a unique dataset containing the price-format decisions of all supermarkets in the U.S. The data contain the format-change decisions of supermarkets in response to a large shock to their local market positions: the entry of Wal-Mart. We exploit the responses of retailers to Wal- Mart entry to infer the cost of changing pricing-formats using a “revealed-preference” argument similar to the spirit of Bresnahan and Reiss (1991). The interaction between retailers and Wal-Mart in each market is modeled as a dynamic game. We find evidence that suggests the entry patterns of WalMart had a significant impact on the costs and incidence of switching pricing strategy. Our results add to the marketing literature on the organization of retail markets, and to a new literature that discusses implications of marketing pricing decisions for macroeconomic studies of price rigidity. More generally, our approach which incorporates long-run dynamic consequences, strategic interaction, and sunk investment costs, outlines how the paradigm of dynamic games may be used to model empirically firms’’positioning decisions in Marketing.

+ Full Paper (PDF)

See: Everyday Low Pricing May Not Be the Best Strategy for Supermarkets

Do We Need Speed Limits on Freeways?

December 13, 2011 Comments off

Do We Need Speed Limits on Freeways? (PDF)

Source:  Stanford University, Department of Economics

When choosing his speed, a driver faces a trade-off between private benefits (time savings) and private costs (fuel cost and own damage and injury). Driving faster also has external costs (pollution, adverse health impacts and injury to other drivers). This paper uses large-scale speed limit increases in the western United States in 1987 and 1996 to address three related questions. First, do the social benefits of raising speed limits exceed the social (private plus external) costs? Second, do the private benefits of driving faster as a result of higher speed limits exceed the private costs? Third, could completely eliminating speed limits improve efficiency? I find that a 10 mph speed limit increase on highways leads to a 3-4 mph increase in travel speed, 9-15% more accidents, 34-60% more fatal accidents, and elevated pollutant concentrations of 14-25% (carbon monoxide), 9-16% (nitrogen oxides), 1-11% (ozone) and 9% higher fetal death rates around the affected freeways. I use these estimates to calculate private and external benefits and costs, and find that the social costs of speed limit increases are three to ten times larger than the social benefits. In contrast, many individual drivers would enjoy a net private benefit from driving faster. Privately, a value of a statistical life (VSL) of $6.0 million or less justifies driving faster, but the social planner’s VSL would have to be below $0.9 million to justify higher speed limits. The substantial difference between private and social optimal speed choices provides a strong rationale for having speed limits. Although speed limits are blunt instruments that differ from an ideal Pigovian tax on speed, it is highly unlikely that any hidden administrative costs or unforeseen behavioral adjustments could make eliminating speed limits an efficiency-improving proposition.

Advertising Effects in Presidential Elections

November 28, 2011 Comments off

Advertising Effects in Presidential Elections
Source: Stanford Graduate School of Business Research Paper

We estimate advertising effects in the context of presidential elections. This setting overcomes many data challenges in previous advertising studies, while arguably providing one of the most interesting empirical settings to study advertising’s effects. The four-year presidential election cycle facilitates measurement in two ways. First, the gap between elections depreciates past advertising stocks such that large advertising investments are concentrated within relatively short periods. Second, the lack of political advertising between elections allows lagged advertising prices to serve as instruments that are safely independent of candidates’ current advertising choices. To further aid estimation, the winner-take-all nature of the electoral college generates broad variation in advertising levels across states. We analyze the data using an aggregate discrete choice approach with extensive fixed effects at the party-market level to control for unobservable cross-sectional factors that might be correlated with advertising, outcomes, and instruments. The results indicate significant positive effects of advertising exposures for the 2000 and 2004 general elections. Advertising elasticities are smaller than are typical for branded goods, yet significant enough to shift election outcomes. For example, if advertising were set to zero and all other factors held constant, three states’ electoral votes would have changed parties in 2000, leading to a different president.

+ Full Paper (PDF)

See: TV Ads Do Influence Consumers — In Elections and Beyond (Stanford Graduate School of Business)

2011 Corporate Board of Directors Survey

September 19, 2011 Comments off

2011 Corporate Board of Directors Survey
Source: Stanford Graduate School of Business (Corporate Governance Research Program)

A new survey from Stanford University’s Rock Center for Corporate Governance and Heidrick & Struggles has uncovered surprises about who makes the best board directors: it’s not necessarily the current CEOs that most companies seek out.

“The popular consensus is that active CEOs make the best board members because of their current strategic and leadership experience,” says David Larcker, professor at the Stanford Graduate School of Business. In the 2011 Corporate Board of Directors Survey, when asked about potential problems a full 87 percent said that active CEOs are too busy with their own companies to be effective directors. A third of the respondents said that active CEOs were “too bossy/used to having their own way.”
“It’s great to have sitting CEOs on a board, but companies need to be aware of the costs associated with having them,” says Stephen A. Miles, Vice Chairman at leadership advisory firm Heidrick & Struggles. “Because active CEOs are so busy, they might be unavailable during a crisis or have to cancel meeting attendance at the last minute. They also have less time to review materials. For some, the demands of their full-time job make it hard for them to consistently be as engaged as they need to be.”

Analyzing responses from 163 directors of public and private companies across North America, the 2011 Corporate Board of Directors Survey reveals how directors think about the composition of the board and the effectiveness of various types of board members.

Related Closer Looks

- Are Current CEOs the Best Board Members? (PDFPDF icon)
- Scarlet Letter: Are the CEOs and Directors of Failed Companies “Tainted”? (PDFPDF icon)

Growth in Data Center Electricity Use 2005 to 2010

August 4, 2011 Comments off

Growth in Data Center Electricity Use 2005 to 2010 (PDF)
Source: Stanford University (Jonathan G. Koomey, Ph.D. Consulting Professor)

Key findings from the new study are as follows:

  • Assuming that the midpoint between the Upper and Lower bound cases accurately reflects the history, electricity used by data centers worldwide increased by about 56% from 2005 to 2010 instead of doubling (as it did from 2000 to 2005), while in the US it increased by about 36% instead of doubling.
  • Electricity used in global data centers in 2010 likely accounted for between 1.1% and 1.5% of total electricity use, respectively. For the US that number was between 1.7 and 2.2%.
  • Electricity used in US data centers in 2010 was significantly lower than predicted by the EPA’s 2007 report to Congress on data centers. That result reflected this study’s reduced electricity growth rates compared to earlier estimates, which were driven mainly by a lower server installed base than was earlier predicted rather than the efficiency improvements anticipated in the report to Congress.
  • While Google is a high profile user of computer servers, less than 1% of electricity used by data centers worldwide was attributable to that company’s data center operations.

Do Proxy Advisory Services Like ISS Create Shareholder Value?

May 19, 2011 Comments off

Do Proxy Advisory Services Like ISS Create Shareholder Value? (PDF)
Source: Stanford Graduate School of Business

Many institutional investors rely on a proxy advisory firm to assist them in voting company proxies and fulfilling the fiduciary responsibility they have to vote in the interest of beneficial shareholders.1 The largest proxy advisory firm is Institutional Shareholder Services (ISS) whose clients have about $25 trillion in assets under management.

There are potential benefits to obtaining voting recommendations from a proxy advisory firm. These firms provide an examination of corporate governance issues (such as director elections, equity compensation plans, structural changes to the board, and charter and bylaw amendments) that might be beyond the expertise of certain investors. This is particularly true for funds that lack staff resources to dedicate to individual proxy items across all companies they are invested in. The recommendations of proxy advisory firms also provide a second opinion to large institutions that do their own proxy analysis.

At the same time, there are potential drawbacks to relying on proxy advisory firms. First, these firms seem to take a somewhat inflexible approach when evaluating governance matters that does not take into account the unique situation, objectives, or business model of the companies they review. As such, their recommendations might reflect a one-size-fits-all approach to governance and the propagation of “best practices” that are typically not supported by the research literature. Second, proxy advisory firms may not have sufficient expert staff to evaluate all items subject to shareholder approval, particularly complex issues such as equity compensation plans, say on pay, and proposed acquisitions.2 Finally, to our knowledge, there is no rigorous evidence that proxy advisory firm recommendations are correct or that they increase shareholder value. To this end, some corporate directors suggest that their recommendations are arbitrary and not in the interest of shareholders.

Patterns of Temporal Variation in Online Media

April 23, 2011 Comments off

Patterns of Temporal Variation in Online Media (PDF)
Source:

Online content exhibits rich temporal dynamics, and diverse realtime user generated content further intensifies this process. However, temporal patterns by which online content grows and fades over time, and by which different pieces of content compete for attention remain largely unexplored.

We study temporal patterns associated with online content and how the content’s popularity grows and fades over time. The attention that content receives on the Web varies depending on many factors and occurs on very different time scales and at different resolutions. In order to uncover the temporal dynamics of online content we formulate a time series clustering problem using a similarity metric that is invariant to scaling and shifting. We develop the K-Spectral Centroid (K-SC) clustering algorithm that effectively finds cluster centroids with our similarity measure. By applying an adaptive wavelet-based incremental approach to clustering, we scale K-SC to large data sets.

We demonstrate our approach on two massive datasets: a set of 580 million Tweets, and a set of 170 million blog posts and news media articles. We find that K-SC outperforms the K-means clustering algorithm in finding distinct shapes of time series. Our analysis shows that there are six main temporal shapes of attention of online content. We also present a simple model that reliably predicts the shape of attention by using information about only a small number of participants. Our analyses offer insight into common temporal patterns of the content on theWeb and broaden the understanding of the dynamics of human attention.

Follow

Get every new post delivered to your Inbox.

Join 858 other followers