Pinned Repositories
Age_prediction
principal component analysis (PCA) to perform dimensionality reduction of the given dataset. Then, train a linear regression model on the reduced-dimension dataset to learn their age.
Assistance_Requiring_Activity_Recognition
This repository contains the code for Deep Learning Project -- "Recognizing Assistance Requiring Activity via Deep Learning''
Customer_attrition_prediction
This predictive analysis is vital for the any banking system to revise their business strategy towards customer retention.
Cytoscape_demo
EHR-Data-Embedding-and-Analysis-for-AFib-Detection
Language_Modeling
Next Word and Character prediction using neural network.
Novel_Extention_to_SPIHT_and_SPECK
Compression is useful because it helps to reduce the consumption of expensive resources, such as hard disk space or transmission bandwidth(computing). In order to achieve maximal storage and transmission capabilities, different compression algorithms should be compared in order to find an optimal technique for image compression. DWT(discrete wavelet transform) under goes two types of coding schemes i.e. tree based coding like SPIHT, EZW and block based coding like SPECK, EBCOT etc. for image compression. In this research, we studied the performance of SPIHT(set partitioning in Hierarchical Trees) and SPECK(set partitioning in embedded block), then we tried a novel extension to these popular schemes. We report results from our comparative study of different lossy image coding using gray scale and color image; PSNR(Peak Signal to Noise Ratio) and SSIM(Structural similarity) as traditional objective picture quality measures, the Algorithm Complexity in Terms of encoding/decoding Time and Memory Requirement. The objective of this work is to provide a quantitative and qualitative comparison of lossy image coding focusing on different bit rate.
SmartFS-an-efficient-File-System
descriptionSmartFS is an automated file system that automatically upload unused files to a remote drive to save storage.
Traffic-and-Weather-data-mining-and-modeling-for-Accident-prediction
Road safety is one of the top most priority of every government and individual. Government spends billions of dollars on making great roadway infrastructure and safety certification of the vehicles, so that the lives of the people on the road will be safer. However, there are still a large number of fatal accidents occurring on the road. In the year 2018 alone, the number of fatalities on the road has increased upto 1.28 million [1]. Predicting accidents have thus become one of the widely researched topics which could be used by different agencies for optimizing traffic conditions (e.g. adding more lanes in one direction and reversing it when the condition changes), provide a dynamic route to riders using GPS and improving overall transportation infrastructure. There are a variety of dataset publicly available for this cause such as accident data, traffic event data and weather data. These datasets could be used to prepare useful classification models to predict whether a particular condition is more prone to accidents and drivers must drive with precaution.
UCBox-a-cloud-based-Android-App
Cloud based android app
prustyp's Repositories
prustyp/Cytoscape_demo
prustyp/EHR-Data-Embedding-and-Analysis-for-AFib-Detection
prustyp/SmartFS-an-efficient-File-System
descriptionSmartFS is an automated file system that automatically upload unused files to a remote drive to save storage.
prustyp/Twitter-Sentiment-Analysis-using-Flask-and-Heroku
prustyp/Assistance_Requiring_Activity_Recognition
This repository contains the code for Deep Learning Project -- "Recognizing Assistance Requiring Activity via Deep Learning''
prustyp/Traffic-and-Weather-data-mining-and-modeling-for-Accident-prediction
Road safety is one of the top most priority of every government and individual. Government spends billions of dollars on making great roadway infrastructure and safety certification of the vehicles, so that the lives of the people on the road will be safer. However, there are still a large number of fatal accidents occurring on the road. In the year 2018 alone, the number of fatalities on the road has increased upto 1.28 million [1]. Predicting accidents have thus become one of the widely researched topics which could be used by different agencies for optimizing traffic conditions (e.g. adding more lanes in one direction and reversing it when the condition changes), provide a dynamic route to riders using GPS and improving overall transportation infrastructure. There are a variety of dataset publicly available for this cause such as accident data, traffic event data and weather data. These datasets could be used to prepare useful classification models to predict whether a particular condition is more prone to accidents and drivers must drive with precaution.
prustyp/UCBox-a-cloud-based-Android-App
Cloud based android app
prustyp/Language_Modeling
Next Word and Character prediction using neural network.
prustyp/Customer_attrition_prediction
This predictive analysis is vital for the any banking system to revise their business strategy towards customer retention.
prustyp/Age_prediction
principal component analysis (PCA) to perform dimensionality reduction of the given dataset. Then, train a linear regression model on the reduced-dimension dataset to learn their age.
prustyp/Novel_Extention_to_SPIHT_and_SPECK
Compression is useful because it helps to reduce the consumption of expensive resources, such as hard disk space or transmission bandwidth(computing). In order to achieve maximal storage and transmission capabilities, different compression algorithms should be compared in order to find an optimal technique for image compression. DWT(discrete wavelet transform) under goes two types of coding schemes i.e. tree based coding like SPIHT, EZW and block based coding like SPECK, EBCOT etc. for image compression. In this research, we studied the performance of SPIHT(set partitioning in Hierarchical Trees) and SPECK(set partitioning in embedded block), then we tried a novel extension to these popular schemes. We report results from our comparative study of different lossy image coding using gray scale and color image; PSNR(Peak Signal to Noise Ratio) and SSIM(Structural similarity) as traditional objective picture quality measures, the Algorithm Complexity in Terms of encoding/decoding Time and Memory Requirement. The objective of this work is to provide a quantitative and qualitative comparison of lossy image coding focusing on different bit rate.