Machine Learning Baskarya🤖

Description

Machine learning make recomendation system based by visual similarity.

Method

Created machine learning model for batik motif classification using a pre trained model from MobileNetV2. Then we take the layer before the output layer to extract images from each batik and look for similarities using cosine similarity to make recommendations.

Tools

• 
• 
• 
• 
• 
• 
• 

Dataset

This dataset is a collection of images collected from various sources.

Sources

• Indonesian Batik Motifs
• Indonesian Batik Motifs (Corak App)
• Dataset-Motif-Batik
• Google Images

Data Preparation

The process of preparing the dataset involved:

1. Collection: Gathering images from multiple sources, ensuring a wide range of samples.

2. Selection: Choosing images based on quality, relevance, and adherence to project criteria.

3. Combination: Merging selected images into a cohesive dataset.

Dataset Result : Dataset

Deployment

We using Flask to deploy machine learning model. Click here to see documentation.

References

• Medium Article: "Image-Based Product Recommendation System"