My experiments and learning towards machine learning
- seperate1.ipynb : When we have to sort through 100's or 1000's pf images in order to keep some images and delete useless images. It becomes pain in itself.
- What i did was created a folder with all the images of whatsapp/ any source.
- Passed them throught an pretrained alexnet net in order to get 512 vector representation of images.
- Applied the K nearest neighbours for the images. With varying values of k such as 2,3,4,5.
Result: I was able to seperate the photographs of people 1 cluster, notes photos 2nd cluster, meme or fun cluster images recieved on whatsapp 3rd cluster. It helped reduce my time to save only relevent images out of all the mixture of images. Save time. List of algorithms
- [ ] Linear Regression
- [X] Logistic Regression
- [ ] Decision Tree
- [ ] SVM
- [ ] Naive Bayes
- [ ] kNN
- [ ] K-Means
- [x] Random Forest
- [ ] Dimensionality Reduction Algorithms
- [ ] Gradient Boosting algorithms
- [ ] GBM
- [ ] XGBoost
- [ ] LightGBM
- [ ] CatBoost
Logistic regression aplications: [] Fraud detection in Credit card [] Email spam or ham [] Sentiment Analysis in Twitter analysis [] Image segmentation, recognition, and classification – X-rays, Scans [] Object detection through video [] Handwriting recognition [] Disease prediction – Diabetes, Cancer, Parkinson etc…