The objective of this machine learning Python application is to identify the transmission of a car by looking at the brand, the model, the year, the type of fuel, and several other features characterizing automobiles in the Indian used car market which is expected to reach USD 74.70 billion in 2027 according to the India Used Car Market report published by Modor Intelligence in 2021.
This is a case study of a supervised learning where the algorithm is first trained on a set of cars, and then used to classify new data.
NB: We have not taken into account any car transmission types other than Manual and Automatic.
In this project, the following methods were used for classification:
- Scikit-learn neural network Multi-layer Perceptron classifier.
- Single Layer Neural Network algorithm built from scratch.
- Scikit-learn Categorical Bayesian Classifier
- Categorical Bayesian Classifier built from scratch.
We look at how these methods perform with our data, interpret the results based on the target and finally investigate the limitations on accuracy posed by the dataset itself if any.