Machine Learning

The data included here within is the associated model training results from the correlated paper "Distribution-Driven Augmentation of Real-World Datasets for Improved Cancer Diagnostics With Machine Learning". This paper focuses on using kernel density estimators to curate datasets by balancing classes and filling missing null values though synthetically generated data. Additionally, this manuscript proposes a technique for joining distinct datasets to train a model with necessary features from multiple different datasets as a type of transfer-learning.

The growing antenna array scale, the uncorrelated fadings between downlink and uplink of frequency division duplex (FDD) or analog beamforming design increases the difficulty of channel sounding or estimation. Non-wireless channel detection or beam weight prediction method is a promising solution to help obtain timely and accurate wireless channel state. Furthermore, beamforming can be enhanced by the powerful sensing capability of cameras.

The process of dataset generation comprises three integral components: "Account Profiles," responsible for creating detailed account representations; "Transaction Generation," which simulates a diverse range of financial transactions; and the "Generation of Fraud Scenarios," which introduces predefined templates for identifying potential fraudulent transactions based on various criteria. Together, these components collaboratively construct a dynamic and realistic dataset, mirroring real-world financial systems.

The dataset consists of measurements of four different stages of degradation in low-voltage contactors used for industrial purposes. The measurements were obtained with fiber Bragg grating (FBG) sensors that detect the dynamic deformation generated in switching under different internal components. The measurements were processed and features from PSD, FFT and TSFEL python library were extracted. The features of PSD and FFT were acquired in 40 sliding windows of 50Hz from the signal.