sirilsam77 Goto Github PK

amazon-fashion-recommendation-system

breast-cancer-diagnosis-using-logistic-regression-using-l2-regularization

A Logistic regression model is built using SGD algorithm for weight and bias update and the final model is compared with Inbuilt function

car-and-swimming-pool-detection-using-satellite-imagery

cervical-cancer-image-reflection-removal

cnn-on-cifar-10-dataset

A test accuracy of 90.04% was acheived by training the densenet model from scratch.There are no dropouts,no fully connected layers used.There are only 0.8 million parameters in total.

cnn-on-mnist

dash.js

A reference client implementation for the playback of MPEG DASH via Javascript and compliant browsers.

data-visualization-and-analysis-on-cancer-survival-patients

decision-trees-on-donorschoose

eyes-face-car-and-pedestrian-detection-using-haar-cascade

facebook-friend-recommendation-challenge-using-graph-link

human-activity-recognition

Human activity Recognition(Classification) is done on data collected using Gyroscope and Accelerometer using deep learning model without any feature engineering it gave a Test accuracy of 93.11% and Machine learning was applied on Feature engineered features which gave an accuracy of 96%.

kmeans-dbscan-on-donorschoose

knn-on-donorschoose

lancovernet-multiclass-segmentation-on-sentinel-2-imagery

Multiclass segmentation of Sentinel 2 data on Africa using ResnetUnet Architecture

logistic-regression-on-donorschoose

lstm-on-donorschoose

mlhub-tutorials

Tutorials to access Radiant MLHub Training Datasets

mnist-mlp

multiclass-segmentation-using-resnet-unet-oon-landcovernet-dataset

Achieved a jaccard index of 0.75 with 100 images.LandCoverNet is a global annual land cover classification training dataset with labels for the multi-spectral satellite imagery from Sentinel-2 mission in 2018. Version 1.0 of the dataset contains data across Africa, which accounts for ~1/5 of the global dataset. Each pixel is identified as one of the seven land cover classes based on its annual time series. These classes are water, natural bare ground, artificial bare ground, woody vegetation, cultivated vegetation, (semi) natural vegetation, and permanent snow/ice. There are a total of 1980 image chips of 256 x 256 pixels in V1.0 spanning 66 tiles of Sentinel-2. Each image chip contains temporal observations from Sentinel-2 surface reflectance product (L2A) at 10m spatial resolution and an annual class label, all stored in a raster format (GeoTIFF files).

naive-bayes-on-donorschoose-dataset

netflix-movie-recommendation-system

new-york-city-taxi-demand-prediction

A MACHIINE LEARNING CASE STUDY TO PREDICT DEMAND OF TAXI'S IN NEW YORK CITY WITH A TEST M.A.P.E OF 8.6% OBTAINED AFTER PERFORMING FEATURE ENGINEERING.

personalized-cancer-diagnosis

A MACHINE LEARNING CLASSIFICATION PROBLEM SOLVED USING FEATURE ENGINEERING TO REDUCE LOG LOSS OF MODEL BELOW 1.

python-programming

python code for 14 questions

quora-questions-pair-similarity

A Machine Learning Case Study to predict the similarity between two questions on Quora. This is a binary-class classification problem.

random-forest-and-gbdt-on-donorschoose

stackoverflow-tag-prediction

svm-on-donorschoose

truncated-svd-on-donorschoose