Machine Learning is the study of methods for programming computers to learn. Computers are applied to a wide range of tasks, and for most of these it is relatively easy for programmers to design and implement the necessary software. However, there are many tasks for which this is difficult or impossible. These can be divided into four general categories. First, there are problems for which there exist no human experts. For example, in modern automated manufacturing facilities, there is a need to predict machine failures before they occur by analyzing sensor readings. Because the machines are new, there are no human experts who can be interviewed by a programmer to provide the knowledge necessary to build a computer system. A machine learning system can study recorded data and subsequent machine failures and learn prediction rules. Second, there are problems where human experts exist, but where they are unable to explain their expertise. This is the case in many perceptual tasks, such as speech recognition, hand-writing recognition, and natural language understanding. Virtually all humans exhibit expert-level abilities on these tasks, but none of them can describe the detailed steps that they follow as they perform them. Fortunately, humans can provide machines with examples of the inputs and correct outputs for these tasks, so machine learning algorithms can learn to map the inputs to the outputs. Third, there are problems where phenomena are changing rapidly. In finance, for example, people would like to predict the future behavior of the stock market, of consumer purchases, or of exchange rates. These behaviors change frequently, so that even if a programmer could construct a good predictive computer program, it would need to be rewritten frequently. A learning program can relieve the programmer of this burden by constantly modifying and tuning a set of learned prediction rules. Fourth, there are applications that need to be customized for each computer user separately. Consider, for example, a program to filter unwanted electronic mail messages. Different users will need different filters. It is unreasonable to expect each user to program his or her own rules, and it is infeasible to provide every user with a software engineer to keep the rules up-to-date. A machine learning system can learn which mail messages the user rejects and maintain the filtering rules automatically. Machine learning addresses many of the same research questions as the fields of statistics, data mining, and psychology, but with differences of emphasis. Statistics focuses on understanding the phenomena that have generated the data, often with the goal of testing different hypotheses about those phenomena. Data mining seeks to find patterns in the data that are understandable by people. Psychological studies of human learning aspire to understand the mechanisms underlying the various learning behaviors exhibited by people (concept learning, skill acquisition, strategy change, etc.).

Machine learning

25. Multiple Imputation for Nonresponse in Surveys. By D. B. Rubin. ISBN 0 471 08705 X. Wiley, Chichester, 1987. 258 pp. £30.25.

Multiple Imputation for Nonresponse in Surveys

Labeling training data is increasingly the largest bottleneck in deploying machine learning systems. We present Snorkel, a first-of-its-kind system that enables users to train state-of-the-art models without hand labeling any training data. Instead, users write labeling functions that express arbitrary heuristics, which can have unknown accuracies and correlations. Snorkel denoises their outputs without access to ground truth by incorporating the first end-to-end implementation of our recently proposed machine learning paradigm, data programming. We present a flexible interface layer for writing labeling functions based on our experience over the past year collaborating with companies, agencies, and research labs. In a user study, subject matter experts build models 2.8X faster and increase predictive performance an average 45.5% versus seven hours of hand labeling. We study the modeling tradeoffs in this new setting and propose an optimizer for automating tradeoff decisions that gives up to 1.8X speedup per pipeline execution. In two collaborations, with the U.S. Department of Veterans Affairs and the U.S. Food and Drug Administration, and on four open-source text and image data sets representative of other deployments, Snorkel provides 132% average improvements to predictive performance over prior heuristic approaches and comes within an average 3.60% of the predictive performance of large hand-curated training sets.

/pdf/snorkel-rapid-training-data-creation-with-weak-supervision-15qyryp476.pdf

Snorkel: rapid training data creation with weak supervision

Data collection is a major bottleneck in machine learning and an active research topic in multiple communities. There are largely two reasons data collection has recently become a critical issue. First, as machine learning is becoming more widely-used, we are seeing new applications that do not necessarily have enough labeled data. Second, unlike traditional machine learning, deep learning techniques automatically generate features, which saves feature engineering costs, but in return may require larger amounts of labeled data. Interestingly, recent research in data collection comes not only from the machine learning, natural language, and computer vision communities, but also from the data management community due to the importance of handling large amounts of data. In this survey, we perform a comprehensive study of data collection from a data management point of view. Data collection largely consists of data acquisition, data labeling, and improvement of existing data or models. We provide a research landscape of these operations, provide guidelines on which technique to use when, and identify interesting research challenges. The integration of machine learning and data management for data collection is part of a larger trend of Big data and Artificial Intelligence (AI) integration and opens many opportunities for new research.

A Survey on Data Collection for Machine Learning: A Big Data - AI Integration Perspective

Crowdsourcing has emerged as a novel problem-solving paradigm, which facilitates addressing problems that are hard for computers, e.g., entity resolution and sentiment analysis. However, due to the openness of crowdsourcing, workers may yield low-quality answers, and a redundancy-based method is widely employed, which first assigns each task to multiple workers and then infers the correct answer (called truth) for the task based on the answers of the assigned workers. A fundamental problem in this method is Truth Inference, which decides how to effectively infer the truth. Recently, the database community and data mining community independently study this problem and propose various algorithms. However, these algorithms are not compared extensively under the same framework and it is hard for practitioners to select appropriate algorithms. To alleviate this problem, we provide a detailed survey on 17 existing algorithms and perform a comprehensive evaluation using 5 real datasets. We make all codes and datasets public for future research. Through experiments we find that existing algorithms are not stable across different datasets and there is no algorithm that outperforms others consistently. We believe that the truth inference problem is not fully solved, and identify the limitations of existing algorithms and point out promising research directions.

/pdf/truth-inference-in-crowdsourcing-is-the-problem-solved-523kuaxe14.pdf

Truth inference in crowdsourcing: is the problem solved?

Crowdsourcing refers to solving large problems by involving human workers that solve component sub-problems or tasks. In data crowdsourcing, the problem involves data acquisition, management, and analysis. In this paper, we provide an overview of data crowdsourcing, giving examples of problems that the authors have tackled, and presenting the key design steps involved in implementing a crowdsourced solution. We also discuss some of the open challenges that remain to be solved.

Challenges in Data Crowdsourcing

Recommendation problems with large numbers of discrete items, such as products, webpages, or videos, are ubiquitous in the technology industry. Deep neural networks are being increasingly used for these recommendation problems. These models use embeddings to represent discrete items as continuous vectors, and the vocabulary sizes and embedding dimensions, despite their heavy influence on the model's accuracy, are often manually selected in a heuristical manner. We present Neural Input Search (NIS), a technique for learning the optimal vocabulary sizes and embedding dimensions for categorical features. The goal is to maximize prediction accuracy subject to a constraint on the total memory used by all embeddings. Moreover, we argue that the traditional Single-size Embedding (SE), which uses the same embedding dimension for all values of a feature, suffers from inefficient usage of model capacity and training data. We propose a novel type of embedding, namely Multi-size Embedding (ME), which allows the embedding dimension to vary for different values of the feature. During training we use reinforcement learning to find the optimal vocabulary size for each feature and embedding dimension for each value of the feature. Experimentation on two public recommendation datasets shows that NIS can find significantly better models with much fewer embedding parameters. We also deployed NIS in production to a real world large scale App ranking model in our company's App store, Google Play, resulting in +1.02% App Install with 30% smaller model size.

https://dl.acm.org/doi/pdf/10.1145/3394486.3403288

Neural Input Search for Large Scale Recommendation Models

Worker quality control is a crucial aspect of crowdsourcing systems; typically occupying a large fraction of the time and money invested on crowdsourcing. In this work, we devise techniques to generate confidence intervals for worker error rate estimates, thereby enabling a better evaluation of worker quality. We show that our techniques generate correct confidence intervals on a range of real-world datasets, and demonstrate wide applicability by using them to evict poorly performing workers, and provide confidence intervals on the accuracy of the answers.

Evaluating the Crowd with Confidence

http://ilpubs.stanford.edu:8090/1093/1/fp0194-joglekar.pdf

Evaluating the crowd with confidence

Cipherbase is a comprehensive database system that provides strong end-to-end data confidentiality through encryption. Cipherbase is based on a novel architecture that combines an industrial strength database engine (SQL Server) with lightweight processing over encrypted data that is performed in secure hardware. The overall architecture provides significant benefits over the state-of-the-art in terms of security, performance, and functionality. This paper presents a prototype of Cipherbase that uses FPGAs to provide secure processing and describes the system engineering details implemented to achieve competitive performance for transactional workloads. This includes hardware-software co-design issues (e.g. how to best offer parallelism), optimizations to hide the latency between the secure hardware and the main system, and techniques to cope with space inefficiencies. All these optimizations were carefully designed not to affect end-to-end data confidentiality. Our experiments with the TPC-C benchmark show that in the worst case when all data are strongly encrypted, Cipherbase achieves 40% of the throughput of plaintext SQL Server. In more realistic cases, if only critical data such as customer names are encrypted, the Cipherbase throughput is more than 90% of plaintext SQL Server.

https://www.microsoft.com/en-us/research/wp-content/uploads/2016/02/cipherbase.pdf

Manas Joglekar

Papers

Challenges in Data Crowdsourcing

Neural Input Search for Large Scale Recommendation Models

Evaluating the Crowd with Confidence

Evaluating the crowd with confidence

Transaction processing on confidential data using cipherbase