Predicting Diabetes Using a Machine Learning Approach - DZone Big Data


Diabetes is one of deadliest diseases in the world. It is not only a disease but also a creator of different kinds of diseases like heart attack, blindness, kidney diseases, etc. The normal identifying process is that patients need to visit a diagnostic center, consult their doctor, and sit tight for a day or more to get their reports. Moreover, every time they want to get their diagnosis report, they have to waste their money in vain. But with the rise of Machine Learning approaches we have the ability to find a solution to this issue, we have developed a system using data mining which has the ability to predict whether the patient has diabetes or not.

Texas hospital struggles to make IBM's Watson cure cancer


If IBM is looking for a new application for its Watson machine learning tools, it might consider putting health care providers' procurement and systems integration woes ahead of curing cancer. The fall-out from that project has now prompted the resignation of the cancer center's president, Ronald DePinho, the Wall Street Journal reported Thursday. The university recently published an internal audit report into the procurement processes that led it to hand almost $40 million to IBM and over $21 million to PwC for work on the project, almost all of it without board approval. It noted that the scope of its review was limited to contracting and procurement practices and compliance issues, and did not cover project management and system development activities. The audit "should not be interpreted as an opinion on the scientific basis or functional capabilities of the system in its current state," because a separate review of those aspects of the project is being conducted by an external consultant, it said.

Parameterized Exploration Artificial Intelligence

We introduce Parameterized Exploration (PE), a simple family of methods for model-based tuning of the exploration schedule in sequential decision problems. Unlike common heuristics for exploration, our method accounts for the time horizon of the decision problem as well as the agent's current state of knowledge of the dynamics of the decision problem. We show our method as applied to several common exploration techniques has superior performance relative to un-tuned counterparts in Bernoulli and Gaussian multi-armed bandits, contextual bandits, and a Markov decision process based on a mobile health (mHealth) study. We also examine the effects of the accuracy of the estimated dynamics model on the performance of PE.

Using Machine Learning to Diagnose Depression - DZone Big Data


Machine learning has been increasingly capable of accurately diagnosing a range of physical and mental health conditions in recent years. I've written previously about algorithms that monitor things like speech to detect the onset of conditions such as Alzheimer's, whilst applications have also used mobile phone data to detect changes in lifestyle and possible depression in individuals. A recent study from the University of Texas at Austin takes a slightly different tact by using AI to spot vulnerability to depression from brain imaging. The researchers worked with a supercomputer to train the algorithm to detect commonalities in MRI scans, genomic data and various other datasets relevant to depression and anxiety. It aims to improve upon previous work by researchers who have studied mental disorders via the relationship between brain function and structure in neuroimaging data.

Thompson Sampling for Noncompliant Bandits Machine Learning

Thompson sampling, a Bayesian method for balancing exploration and exploitation in bandit problems, has theoretical guarantees and exhibits strong empirical performance in many domains. Traditional Thompson sampling, however, assumes perfect compliance, where an agent's chosen action is treated as the implemented action. This article introduces a stochastic noncompliance model that relaxes this assumption. We prove that any noncompliance in a 2-armed Bernoulli bandit increases existing regret bounds. With our noncompliance model, we derive Thompson sampling variants that explicitly handle both observed and latent noncompliance. With extensive empirical analysis, we demonstrate that our algorithms either match or outperform traditional Thompson sampling in both compliant and noncompliant environments.