Datasets for Course Project

Please use this thread to share & discuss datasets for course project.

General sources:

Indian stocks data

Indian Air Quality Data

Indian Covid-19 Dataset

World Covid-19 Dataset

USA Covid-19 Dataset

Megapixels Dataset for Face Detection, GANs, Human Localization

Agriculture based dataset

India Digital Payments UPI

India Consumption of LPG

India Import/Export Crude OIl

US Unemployment Rate Data

India Road accident Data

Data science Jobs Data

H1-b Visa Data

Donald Trump’s Tweets

Hilary Clinton and Trump’s Tweets

Asteroid Dataset

Solar flares Data

Human face generation GANs

F-1 Race Data

Automobile Insurance



Dota 2





I have chosen this kaggle jigsaw puzzle dataset for course project.

Inputs are images like from jigsaw puzzle pieces (9/9 grid cells in our eye). Labels are in the (1,9) array shape containing integers representing the place of 9 grid cells.

I have tried a few of them, but didn’t seem to work.
So, what loss function to use and how to structure the labels?
Should I do one hot encoding to the labels?



I’m going to use a multimodal dataset for Stress and Affect Detection from here. Do you have any resources I can read to learn how to train a CNN using multiple sensor data?

I am using sign language MNIST data to predict sign languages .
In my course project i have used IMFDB dataset to predict age of an movie actors containing images with different dimensions.

Hi! I uploaded this dataset into Kaggle for my course project. You can see it here.. It contains sunny or cloudy images from outdoors.

Hey Everyone, If someone is working on Facial Expression Recognition, I suggest you have a look at this Data set. While I was working on FERC13 Dataset, I didn’t found it beginner-friendly. so I took to myself and created an easier and cleaner version of it, so any beginner in the data science field can work on very easily. The Dataset contains train set, test set, and labels.txt, and I tried my best to make it a very easy to go dataset, without a hassle to convert any string of pixels to images.

1 Like