Data Science and Visualization Application

An Application for Data Science and Visualization made by students of the Master of Automation and IT program 2022-23. To offer a hands-on, visual, user-friendly experience with data, the application draws on a range of fields like data manipulation, regression, classification, and artificial intelligence.

Features of the Application

  • Data Selection :

    • Provides the user with the option to upload their own dataset or work with one of the two pre-defined datasets
    • Works on Comma Seperated Value (CSV) files
    • Automatic delimiter detection for the CSV files

  • Data Preview :

    • Provides the user with a preview of the selected dataset in the form of a dataframe
    • Displays statistical information for the selected dataset (along with information about the scientific background for the pre-defined datasets)
    • Provides the user with additional features to remove NaN values from the dataset (if any), and reset the index

  • Data Smoothing, Interpolation, and Outlier Recognition :

    • Allows the user to smoothen, perform interpolation, and perform outlier recognition on the used dataset using various methods
    • Allows the user to tune the parameters regarding each method and analyze the corresponding effects
    • Provides the user with information about each method and the corresponding parameters
    • Allows the user to download filtered datasets after outlier recognition

  • AI Based Classification and Regression :

    • Allows the user to perform classification on classification-type datasets
    • Allows the user to perform regression on time-series type datasets
    • Provides the user with information about each method and the corresponding parameters
    • Provides the user with necessary textual and graphical results
    • Allows the user to save the models after training
    • Allows the user to upload pre-trained models to test the accuracy

  • ML Based Classification and Regression :

    • Allows the user to perform classification on classification-type datasets
    • Allows the user to perform regression on time-series type datasets
    • Provides the user with information about each method and the corresponding parameters
    • Provides the user with necessary textual and graphical results
    • Allows the user to test the trained models by uploading test datasets

Created as part of Data Science and Visualization Application.