I have developed "tennis elbow" from lugging this book around the past four weeks, but it is worth the pain, the effort, and the aspirin. It is also worth the (relatively speaking) bargain price. Including appendixes, this book contains 894 pages of text. The entire panorama of the neural sciences is surveyed and examined, and it is comprehensive in its scope, from genomes to social behaviors. The editors explicitly state that the book is designed as "an introductory text for students of biology, behavior, and medicine," but it is hard to imagine any audience, interested in any fragment of neuroscience at any level of sophistication, that would not enjoy this book. The editors have done a masterful job of weaving together the biologic, the behavioral, and the clinical sciences into a single tapestry in which everyone from the molecular biologist to the practicing psychiatrist can find and appreciate his or

Principles of Neural Science

How can we explain the predictions of a black-box model? In this paper, we use influence functions -- a classic technique from robust statistics -- to trace a model's prediction through the learning algorithm and back to its training data, thereby identifying training points most responsible for a given prediction. To scale up influence functions to modern machine learning settings, we develop a simple, efficient implementation that requires only oracle access to gradients and Hessian-vector products. We show that even on non-convex and non-differentiable models where the theory breaks down, approximations to influence functions can still provide valuable information. On linear models and convolutional neural networks, we demonstrate that influence functions are useful for multiple purposes: understanding model behavior, debugging models, detecting dataset errors, and even creating visually-indistinguishable training-set attacks.

Understanding Black-box Predictions via Influence Functions

How can we explain the predictions of a black-box model? In this paper, we use influence functions — a classic technique from robust statistics — to trace a model's prediction through the learning algorithm and back to its training data, thereby identifying training points most responsible for a given prediction. To scale up influence functions to modern machine learning settings, we develop a simple, efficient implementation that requires only oracle access to gradients and Hessian-vector products. We show that even on non-convex and non-differentiable models where the theory breaks down, approximations to influence functions can still provide valuable information. On linear models and convolutional neural networks, we demonstrate that influence functions are useful for multiple purposes: understanding model behavior, debugging models, detecting dataset errors, and even creating visually-indistinguishable training-set attacks.

/pdf/understanding-black-box-predictions-via-influence-functions-285r2tv1ec.pdf

Understanding black-box predictions via influence functions

As data becomes the fuel driving technological and economic growth, a fundamental challenge is how to quantify the value of data in algorithmic predictions and decisions. For example, in healthcare and consumer markets, it has been suggested that individuals should be compensated for the data that they generate, but it is not clear what is an equitable valuation for individual data. In this work, we develop a principled framework to address data valuation in the context of supervised machine learning. Given a learning algorithm trained on $n$ data points to produce a predictor, we propose data Shapley as a metric to quantify the value of each training datum to the predictor performance. Data Shapley value uniquely satisfies several natural properties of equitable data valuation. We develop Monte Carlo and gradient-based methods to efficiently estimate data Shapley values in practical settings where complex learning algorithms, including neural networks, are trained on large datasets. In addition to being equitable, extensive experiments across biomedical, image and synthetic data demonstrate that data Shapley has several other benefits: 1) it is more powerful than the popular leave-one-out or leverage score in providing insight on what data is more valuable for a given learning task; 2) low Shapley value data effectively capture outliers and corruptions; 3) high Shapley value data inform what type of new data to acquire to improve the predictor.

Data Shapley: Equitable Valuation of Data for Machine Learning

Medical data is one of the most rewarding and yet most complicated data to analyze. How can healthcare providers use modern data analytics tools and technologies to analyze and create value from complex data? Data analytics, with its promise to efficiently discover valuable pattern by analyzing large amount of unstructured, heterogeneous, non-standard and incomplete healthcare data. It does not only forecast but also helps in decision making and is increasingly noticed as breakthrough in ongoing advancement with the goal is to improve the quality of patient care and reduces the healthcare cost. The aim of this study is to provide a comprehensive and structured overview of extensive research on the advancement of data analytics methods for disease prevention. This review first introduces disease prevention and its challenges followed by traditional prevention methodologies. We summarize state-of-the-art data analytics algorithms used for classification of disease, clustering (unusually high incidence of a particular disease), anomalies detection (detection of disease) and association as well as their respective advantages, drawbacks and guidelines for selection of specific model followed by discussion on recent development and successful application of disease prevention methods. The article concludes with open research challenges and recommendations.

/pdf/big-data-analytics-for-preventive-medicine-2oyqhfvqua.pdf

Big data analytics for preventive medicine

Management of hyperglycemia in hospitalized patients has a significant bearing on outcome, in terms of both morbidity and mortality. However, there are few national assessments of diabetes care during hospitalization which could serve as a baseline for change. This analysis of a large clinical database (74 million unique encounters corresponding to 17 million unique patients) was undertaken to provide such an assessment and to find future directions which might lead to improvements in patient safety. Almost 70,000 inpatient diabetes encounters were identified with sufficient detail for analysis. Multivariable logistic regression was used to fit the relationship between the measurement of HbA1c and early readmission while controlling for covariates such as demographics, severity and type of the disease, and type of admission. Results show that the measurement of HbA1c was performed infrequently (18.4%) in the inpatient setting. The statistical model suggests that the relationship between the probability of readmission and the HbA1c measurement depends on the primary diagnosis. The data suggest further that the greater attention to diabetes reflected in HbA1c determination may improve patient outcomes and lower cost of inpatient care.

/pdf/impact-of-hba1c-measurement-on-hospital-readmission-rates-1f0b09oa93.pdf

Impact of HbA1c Measurement on Hospital Readmission Rates: Analysis of 70,000 Clinical Database Patient Records

In many public and private Cloud systems, users need to specify a limit for the amount of resources (CPU cores and RAM) to provision for their workloads. A job that exceeds its limits might be throttled or killed, resulting in delaying or dropping end-user requests, so human operators naturally err on the side of caution and request a larger limit than the job needs. At scale, this results in massive aggregate resource wastage. To address this, Google uses Autopilot to configure resources automatically, adjusting both the number of concurrent tasks in a job (horizontal scaling) and the CPU/memory limits for individual tasks (vertical scaling). Autopilot walks the same fine line as human operators: its primary goal is to reduce slack - the difference between the limit and the actual resource usage - while minimizing the risk that a task is killed with an out-of-memory (OOM) error or its performance degraded because of CPU throttling. Autopilot uses machine learning algorithms applied to historical data about prior executions of a job, plus a set of finely-tuned heuristics, to walk this line. In practice, Autopiloted jobs have a slack of just 23%, compared with 46% for manually-managed jobs. Additionally, Autopilot reduces the number of jobs severely impacted by OOMs by a factor of 10. Despite its advantages, ensuring that Autopilot was widely adopted took significant effort, including making potential recommendations easily visible to customers who had yet to opt in, automatically migrating certain categories of jobs, and adding support for custom recommenders. At the time of writing, Autopiloted jobs account for over 48% of Google's fleet-wide resource usage.

https://dl.acm.org/doi/pdf/10.1145/3342195.3387524

Autopilot: workload autoscaling at Google

Sphere Support Vector Machines for large classification tasks

The paper introduces a multi-layer multi-column model of the cortex that uses four different neuron types and short-term plasticity dynamics. It was designed with details of neuronal connectivity available in the literature and meets these conditions: (1) biologically accurate laminar and columnar flows of activity, (2) normal function of low-threshold spiking and fast spiking neurons, and (3) ability to generate different stages of epileptiform activity. With these characteristics the model allows for modeling lesioned or malformed cortex, i.e. examine properties of developmentally malformed cortex in which the balance between inhibitory neuron subtypes is disturbed.

Simulating vertical and horizontal inhibition with short-term dynamics in a multi-column multi-layer model of neocortex.

A mechanism is provided in a data processing system for corpus quality analysis. The mechanism applies at least one filter to a candidate corpus to determine a degree to which the candidate corpus supplements existing corpora for performing a natural language processing (NLP) operation. Responsive to a determination to add the candidate corpus to the existing corpora based on a result of applying the at least one filter, the mechanism adds the candidate corpus to the existing corpora to form modified corpora. The mechanism performs the NLP operation using the modified corpora.

Beata Strack

Papers

Impact of HbA1c Measurement on Hospital Readmission Rates: Analysis of 70,000 Clinical Database Patient Records

Autopilot: workload autoscaling at Google

Sphere Support Vector Machines for large classification tasks

Simulating vertical and horizontal inhibition with short-term dynamics in a multi-column multi-layer model of neocortex.

Corpus Quality Analysis