Good training data is a prerequisite to develop useful Machine Learning applications. However, in many domains existing data sets cannot be shared due to privacy regulations (e.g., from medical studies). This work investigates a simple yet unconventional approach for anonymized data synthesis to enable third parties to benefit from such anonymized data. We explore the feasibility of learning implicitly from visually unrealistic, task-relevant stimuli, which are synthesized by exciting the neurons of a trained deep neural network. As such, neuronal excitation can be used to generate synthetic stimuli. The stimuli data is used to train new classification models. Furthermore, we extend this framework to inhibit representations that are associated with specific individuals. Extensive comparative empirical investigation shows that different algorithms trained on the stimuli are able to generalize successfully on the same task as the original model.

Sleep apnea is a common yet severely under-diagnosed sleep related disorder. Unattended sleep monitoring at home with low-cost sensors can be leveraged for condition detection, and Machine Learning offers a generalized solution for this task. However, patient characteristics, lack of sufficient training data, and other factors can imply a domain shift between training and end-user data and reduced task performance. In this work, we address this issue with the aim to achieve personalization based on the patient’s needs. We present an unsupervised domain adaptation (UDA) solution with the constraint that labeled source data are not directly available. Instead, a classifier trained on the source data is provided. Our solution iteratively labels target data sub-regions based on classifier beliefs, and trains new classifiers from the expanding dataset. Experiments with sleep monitoring datasets and various sensors show that our solution outperforms the classifier trained on the source domain, with a kappa coefficient improvement from 0.012 to 0.242. Additionally, we apply our solution to digit classification DA between three well-established datasets, to investigate its generalizability, and allow for related work comparisons. Even without direct access to the source data, it outperforms several well-established UDA methods in these datasets.

Good training data is a prerequisite to develop useful Machine Learning applications. However, in many domains existing data sets cannot be shared due to privacy regulations (e.g., from medical studies). This work investigates a simple yet unconventional approach for anonymized data synthesis to enable third parties to benefit from such anonymized data. We explore the feasibility of learning implicitly from visually unrealistic, task-relevant stimuli, which are synthesized by exciting the neurons of a trained deep neural network. As such, neuronal excitation can be used to generate synthetic stimuli. The stimuli data is used to train new classification models. Furthermore, we extend this framework to inhibit representations that are associated with specific individuals. We use sleep monitoring data from both an open and a large closed clinical study, and Electroencephalogram sleep stage classification data, to evaluate whether (1) end-users can create and successfully use customized classification models, and (2) the identity of participants in the study is protected. Extensive comparative empirical investigation shows that different algorithms trained on the stimuli are able to generalize successfully on the same task as the original model. Architectural and algorithmic similarity between new and original models play an important role in performance. For similar architectures, the performance is close to that of using the original data (e.g., Accuracy difference of 0.56%-3.82%, Kappa coefficient difference of 0.02-0.08). Further experiments show that the stimuli can provide state-ofthe-art resilience against adversarial association and membership inference attacks.

There are large skill differences among software developers, and clients and managers will benefit from being able to identify those with better skill. This study examines the relations between low effort estimates, and other commonly used skill indicators, and measured programming skill. One hundred and four professional software developers were recruited. After skill-related information was collected, they were asked to estimate the effort for four larger and five smaller programming tasks. Finally, they completed a programming skill test. The lowest and most over-optimistic effort estimates for the larger tasks were given by those with the lowest programming skill, which is in accordance with the well-known Dunning-Kruger effect. For the smaller tasks, however, those with the lowest programming skill had the highest and most over-pessimistic estimates. The other programming skill indicators, such as length of experience, company assessed skill and self-assessed skill, were only moderately correlated with measured skill and not particularly useful in guiding developer skill identification. A practical implication is that for larger and more complex tasks, the use of low effort estimates and commonly used skill indicators as selection criteria leads to a substantial risk of selecting among the least skilled developers.

Supervised machine learning applications in the health domain often face the problem of insufficient training datasets. The quantity of labelled data is small due to privacy concerns and the cost of data acquisition and labelling by a medical expert. Furthermore, it is quite common that collected data are unbalanced and getting enough data to personalize models for individuals is very expensive or even infeasible. This paper addresses these problems by (1) designing a recurrent Generative Adversarial Network to generate realistic synthetic data and to augment the original dataset, (2) enabling the generation of balanced datasets based on heavily unbalanced dataset, and (3) to control the data generation in such a way that the generated data resembles data from specific individuals. We apply these solutions for sleep apnea detection and study in the evaluation the performance of four well-known techniques, i.e., K-Nearest Neighbour, Random Forest, Multi-Layer Perceptron, and Support Vector Machine. All classifiers exhibit in the experiments a consistent increase in sensitivity and a kappa statistic increase by between 0.007 and 0.182.

Abstract Summary: CGH-Explorer is a program for visualization and statistical analysis of microarray-based comparative genomic hybridization (array-CGH) data. The program has preprocessing facilities, tools for graphical exploration of individual arrays or groups of arrays, and tools for statistical identification of regions of amplification and deletion. Availability: The program is available as Java class files that runs on any platform with the Java 2 runtime environment (J2SE JRE) installed, and as a Windows executable. Java source files are also available. See http://www.ifi.uio.no/bioinf/Papers/CGH/ Contact: [email protected]

A pacemaker neuron with the following properties is considered. After a firing, the membrane potential is reset to a constant value from which it increases to the firing threshold during a time t0. The neuron receives strong synaptic input producing postsynaptic potentials (PSPs) which change the membrane potential to the reversal potential of the synapse, from which level the potential increases to the firing threshold during a time t1. Provided that interarrival times for the PSPs are independent and identically distributed, successive interspike intervals in this class of model neurons can be described by a regenerative stochastic process simple enough to allow the derivation of tractable expressions for the limiting distribution of the interspike intervals, including a simple expression for the mean firing rate. A central limit theorem for the partial sums of interspike intervals can also be proved. This class of models is a generalization of a model of the crayfish's stretch receptors, a commonly used neurophysiological system. In two examples the model is studied under varying temporal patterns for the PSPs to illustrate, respectively, phase locking and certain principles of summation of excitation and inhibition.