Classification of Caffeine Products

Project Description:

To create a classification model which can classify various caffeine products into numerous coffee types from its FTIR data. To train a validated classification model on the Infrared Spectroscopy of various caffeine products from different classes of coffee types

Project Objectives:

1. A quick fast and effecient test for product classification as well as quality control of the product.
2. Provide a way to classify the product without causing any damage to the product.
3. This is where The FTIR Spectroscopes step in. They provide a very detailed chemical composition of the product at the same time uses a very small amount of the product, but also does not cause any damage to it.

Project Pipeline:

1. Find IR Spectroscopy values for various classes of caffeine products.
2. Apply FTIR pre-processing algorithms such as smootening, interpolatation, SNV normalization to eliminate noise, normalize the intensities and wavelnegths.
3. Validate Dimensionalty Reduction and Data split
4. Validate multiple Classification models for optimum parameters to achieve maximum accuracy and finally build a classification model to classify caffeine products.
5. Visualize FTIR Spectroscopes and Final Model.

For tech stack used:

pip install requirements.txt