Parkinson's Disease (PD) is a progressive neurodegenerative disease with multiple motor and non-motor characteristics. PD patients commonly face vocal impairments during the early stages of the disease. So, diagnosis systems based on vocal disorders are at the forefront on recent PD detection studies. Our study proposes two frameworks based on Convolutional Neural Networks to classify Parkinson's Disease (PD) using sets of vocal (speech) features. Although, both frameworks are employed for the combination of various feature sets, they have difference in terms of combining feature sets. While the first framework combines different feature sets before given to 9-layered CNN as inputs, whereas the second framework passes feature sets to the parallel input layers which are directly connected to convolution layers. Thus, deep features from each parallel branch are extracted simultaneously before combining in the merge layer. Proposed models are trained with dataset taken from UCI Machine Learning repository and their performances are validated with Leave-One-Person-Out Cross Validation (LOPO CV). Due to imbalanced class distribution in our data, F-Measure and Matthews Correlation Coefficient metrics are used for the assessment along with accuracy. Experimental results show that the second framework seems to be very promising, since it is able to learn deep features from each feature set via parallel convolution layers. Extracted deep features are not only successful at distinguishing PD patients from healthy individuals but also effective in boosting up the discriminative power of the classifiers.

/pdf/deep-learning-based-parkinson-s-disease-classification-using-q8tvwb0xn2.pdf

Deep Learning-Based Parkinson’s Disease Classification Using Vocal Feature Sets

Imbalance-XGBoost: leveraging weighted and focal losses for binary label-imbalanced classification with XGBoost

Horizontal Federated learning (FL) handles multi-client data that share the same set of features, and vertical FL trains a better predictor that combine all the features from different clients. This paper targets solving vertical FL in an asynchronous fashion, and develops a simple FL method. The new method allows each client to run stochastic gradient algorithms without coordination with other clients, so it is suitable for intermittent connectivity of clients. This method further uses a new technique of perturbed local embedding to ensure data privacy and improve communication efficiency. Theoretically, we present the convergence rate and privacy level of our method for strongly convex, nonconvex and even nonsmooth objectives separately. Empirically, we apply our method to FL on various image and healthcare datasets. The results compare favorably to centralized and synchronous FL methods.

VAFL: a Method of Vertical Asynchronous Federated Learning

Most real-world data are scattered across different companies or government organizations, and cannot be easily integrated under data privacy and related regulations such as the European Union's General Data Protection Regulation (GDPR) and China' Cyber Security Law. Such data islands situation and data privacy & security are two major challenges for applications of artificial intelligence. In this paper, we tackle these challenges and propose a privacy-preserving machine learning model, called Federated Forest, which is a lossless learning model of the traditional random forest method, i.e., achieving the same level of accuracy as the non-privacy-preserving approach. Based on it, we developed a secure cross-regional machine learning system that allows a learning process to be jointly trained over different regions' clients with the same user samples but different attribute sets, processing the data stored in each of them without exchanging their raw data. A novel prediction algorithm was also proposed which could largely reduce the communication overhead. Experiments on both real-world and UCI data sets demonstrate the performance of the Federated Forest is as accurate as the non-federated version. The efficiency and robustness of our proposed system had been verified. Overall, our model is practical, scalable and extensible for real-life tasks.

Federated Forest

Soil compaction has contrasting effect on soil strength (i.e., positive) and vegetation growth (i.e., negative), respectively. Biochar has been utilized mostly in combination with soils in both agricultural fields (i.e., loose soils) and geo-structures (i.e., dense soil slopes, landfill cover) for improving water retention due to its microporous structure. Biochar is also found to be useful to reduce gas permeability in compacted soil recently. However, the efficiency of biochar in reducing gas permeability in loose and dense soils is rarely understood. The objective of this study is to analyze effects of compaction on gas permeability in soil at different degrees of compaction (i.e., 65%, 80% and 95%) and also different biochar amendment contents (0%, 5% and 10%). Another aim is to identify relative significance of parameters (soil suction, water content, biochar content and compaction) in affecting gas permeability. Experiments were conducted before applying k-nearest neighbor (KNN) modeling technique for identifying relative significance of parameters. Biochar was synthesized from a coastal invasive species (water hyacinth), which has relatively no influence on food chain (as unlike in biochar produced from biomass such as rice husk, straw, peanut shell). Based on measurements and KNN modeling, it was found that gas permeability of biochar-amended soil is relatively lower than that of soil without amendment. It was found from KNN model that for denser soils, higher amount of soil suction is mobilized for a significant increase in gas permeability as compared to loose soils. Among all parameters, soil suction is found to be most influential in affecting gas permeability followed by water content and compaction.

Mechanism of biochar soil pore-gas-water interaction: gas properties of biochar-amended sandy soil at different degrees of compaction using KNN modeling

A comparative analysis of speech signal processing algorithms for Parkinson’s disease classification and the use of the tunable Q-factor wavelet transform

Predicting permeability of compacted clay filtrated with landfill leachate by k-Nearest Neighbors modelling method

In this study, base-2 computer arithmetic procedures are developed for the analyzing of fast division algorithms; Non- Restoring Division, SRT division and Division by Multiplication algorithms. For this purpose, those algorithms have been developed in MATLAB software simulation platform to compare the performance of the three division method. Depending on the number of bits used in simulation, memory space (A), the number of transactions (I) and the algorithms executed by the CPU processing time (T) criteria were used in the analyzing. As a result of our comparisons, division by multiplication algorithm had achieved better in those three criterias.

A comparative performance analysis for the computer arithmetic based fast division algorithms

Studies on Parkinson's disease (PD) are becoming very popular on multidisciplinary platforms. The development of predictable telemonitored early detection models has become closely related to many different research areas. The aim of this article is to develop a visual performance test that can examine the effects of Parkinson's disease on the visual cortex, which can be a subtitle scoring test in UPDRS. However, instead of showing random images and asking for discrepancies between them, it is expected that the questions to be asked to patients should be provable in the existing cortex models, should be deduced between the images, and produce a reference threshold value to compare with the practical results. In a developed test, horizontal and vertical motion blur orientation was applied to natural image samples, and then neural outputs were produced by representing three (original-horizontal-vertical) image groups with the Layer 4 (L4) cortex model. This image representation is then compared with a filtering model which is very similar to thalamus' functionality. Thus, the linear problem-solving performance of the L4 cortex model is also addressed in the study. According to the obtained classification results, the L4 model produces high-performance success rates compared to the thalamic model, which shows the adaptation power of the visual cortex on the image pattern differences. In future studies, developed motion-based visual tests are planned to be applied to PD patient groups/controls, and their performances with mathematical threshold values will be examined.

/pdf/cortical-visual-performance-test-setup-for-parkinson-s-3wf9utz4tc.pdf

Cortical Visual Performance Test Setup for Parkinson's Disease Based on Motion Blur Orientation.

The Graphical User Interfaces (GUIs) of web applications include visuals and designs that allow users to interact with machines. Once the GUI design is done, it is necessary to generate its GUI code. However, the GUI code generation process is highly time consuming as well as highly dependent on software developers. Therefore, the development of automatic GUI code generating systems is of great importance recently. In this study, a GUI code generating system for web sites is designed using the Deep Learning (DL) approach. The dataset including “coordinate, width, height and type” of GUI objects is created using 7500 webpages. The created dataset is applied to the proposed system in order to detect objects in the GUI image and generate DSL mark-up code. Experiments were carried out to analyze the effectiveness of the proposed system and the performance evaluations were made.

M. Erdem Isenkul

Papers

A comparative analysis of speech signal processing algorithms for Parkinson’s disease classification and the use of the tunable Q-factor wavelet transform

Predicting permeability of compacted clay filtrated with landfill leachate by k-Nearest Neighbors modelling method

A comparative performance analysis for the computer arithmetic based fast division algorithms

Cortical Visual Performance Test Setup for Parkinson's Disease Based on Motion Blur Orientation.

A Deep Learning Based Object Detection System for User Interface Code Generation