Scribd
Upload
4 views2 pages

Mla1 2024

This document outlines the instructions for a Machine Learning assignment focused on multimodal data analysis, specifically involving audio, image, and time-series data. Students must use Google Colab or Jupyter Notebooks, adhere to strict submission guidelines, and utilize only specified libraries. The assignment consists of beginner, intermediate, and advanced tasks involving EEG data analysis, including random number generation, normalization, exploratory data analysis, and feature extraction.

Uploaded by

Aniket Dwivedi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
4 views2 pages

Mla1 2024

This document outlines the instructions for a Machine Learning assignment focused on multimodal data analysis, specifically involving audio, image, and time-series data. Students must use Google Colab or Jupyter Notebooks, adhere to strict submission guidelines, and utilize only specified libraries. The assignment consists of beginner, intermediate, and advanced tasks involving EEG data analysis, including random number generation, normalization, exploratory data analysis, and feature extraction.

Uploaded by

Aniket Dwivedi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Machine Learning Assignment - 1

Instructions.
• This is an easy assignment which delves into the realm of multimodal data analysis, focusing on
the analysis of audio, image and time-series data.
• Use [Google Colab / Jupyter Notebooks] only for this assignment. Not doing so will lead to (0
Marks) in complete assignment.
• Each question has its own instruction, follow precisely. If not (0 Marks) will be awarded.
• Submission Guidelines
– Submit only a single [MLA1 *.ipynb] where * is your roll number.
– For eg: if your roll number is [MT10002], your file name should be [MLA1 MT10002.ipynb]
• Not following naming convention will lead to (0 Marks) , whatever the reason you present.
• You are allowed to use only [pandas, numpy, PIL, matplotlib, scipy ] for this assignment. Again
not following the same will lead to (0 Marks) for entire assignment.
• If you are not able to explain in viva, irrespective of your answer correctnes you will get (0
Marks)
• If there is in front of question you have to implement it completely from scratch and if it is
you can use functions from above mentioned packages.

[ 30 Marks ]
Dataset download link.
Dataset Description: A dataset comprising electroencephalography (EEG) signals, corresponding visual stimuli,
and auditory data from human participants is provided for analysis. EEG data were collected at a sampling rate
of 128 kHz using a five-electrode configuration (AF3, AF4, T7, T8, Pz) while participants viewed and cognitively
processed visual stimuli for three-second intervals. The original EEG data, in .eeg format, has been converted to .csv
for convenience and is accompanied by the respective image and audio files.
• Beginner
1. (2 Marks) Make a random number generator which takes input as size and generate random numbers.
Using this generate and plot random numbers of size:
(a) 1×n
(b) n×n
2. (1 Marks) Using above implemented random number generator select 12 samples from the EEG data
and print there metadata.
3. (3 Marks) Implement a normalization function which can normalize given data in range [-x, x ].
Normalize and plot 4 randomly selected EEG signals.
4. (4 Marks) Use selected EEG signal from previous question and perform on:
• Image Data:
(a) Reshape image to (3,224,224).
(b) Convert to Black&White.
(c) Plot histogram of the image (Color & BW).
(d) Calculate CenterOfMass for all images.
• Audio Data:
(a) Resample audio to 16000.
(b) Convert to MelSpectrogram.
(c) Plot time-domain & frequency-domain signal.
(d) Calculate ZeroCrossingRate for all time-domain signals.
5. (3 Marks) Calculate and plot PowerSpectralDensity (PSD) of randomly selected 4 EEG signals.

1
• Intermediate
1. (4 Marks) Calculate and plot Cross-Correlation and Auto-Correlation of randomly selected 4 EEG
signals.
2. (4 Marks) Perform exploratory data analysis (EDA). (Visualization, statistical analysis, class imbal-
ance, features visualization, outliers detection etc.)
(a) Image data.
(b) Audio data.
3. (4 Marks) Select features from your EDA and perform principal component analysis (PCA) on image
and audio feature data separately. Plot the feature map before and after performing PCA.
• Advanced
1. (5 Marks) Randomly select 120 EEG samples and form a feature-bank for image, audio, and EEG data
separately. Write a pipeline for selecting data, making train-test splits and training-testing classification
using linear regression as classification algorithm.

You might also like