Machine Learning Mini Project

Mini Project on Image Classification using Machine Learning

The following steps were followed to build the classification model to differentiate between the images of bikes and cars:

1. Data Gathering: The data was gathered using Bing Image Downloader

2. Data Preprocessing: The data was preprocessed by-

   2.1. Scaling down the images to a particular size.

   2.2. Flattening the images.

3. Support Vector Machine (SVM) algorithm was implemented for classification of the images.

4. GridSearchCV was used to find the best parameters.

5. The model was evaluated using Confusion Matrix and Accuracy Score and was found to be 93% accurate.

Finally, the user is asked to feed the url link of new images for the pupose of predicting if the images pertain to apple or lemon.

Here's the link to the web page:

https://imag-classifier.herokuapp.com/