Pinned Repositories
Acute-Lymphocytic-Leukemia-ALL-Cell-Classification
Due to morphological similarity at the microscopic level, making an accurate and time sensitive distinction between blood cells affected by Acute Lymphocytic Leukemia (ALL) and their healthy counterparts calls for the usage of machine learning architectures. However, three of the most common models, VGG, ResNet, and Inception, each come with their own set of flaws with room for improvement which demands the need for a superior model. ALLNet, the proposed hybrid convolutional neural network architecture, consists of a combination of the VGG, ResNet, and Inception models. The ALL Challenge dataset of ISBI 2019 contains 10,691 images of white blood cells which were used to train and test the models, 7,272 of cells with ALL and 3,419 of those that were healthy. Of the images, 60% was used to train the model, 20% for the cross validation set, and 20% for the test set. ALLNet outperformed the VGG, Resnet, and the Inception models across the board, achieving an accuracy of 92.6567%, a sensitivity of 95.5304%, a specificity of 85.9155%, an AUC score of 0.966347, and an F1 score of 0.94803 in the cross validation set. In the test set, the ALLNet achieved an accuracy of 92.0991%, a sensitivity of 96.5446%, a specificity of 82.8035%, an AUC score of 0.959972, and an F1 score of 0.942963. The utilization of ALLNet in the clinical workspace can better treat the thousands of people suffering with ALL across the world, many of whom are children.
BostonHousingPricesPredictions
ML model that predicts Boston Housing prices based on data from the boston_housing dataset from keras.
BrainTumorSegmentationAI
This project uses the 2D U-Net architecture to make a machine learning model to segment a brain tumor in an MRI scan of brain. This project was done in Python and used various libraries and frameworks including Tensorflow (keras), numpy, pandas, and more. This program creates 4 different U-Net models, each with different depth (2, 3, 4, and 5). I then train these on the training set and evaluate them on the validation set.
deepcgh3
DeepCGH3
Exo-Potato
Exo-Potato is a software for deep space exoplanet hunting that uses machine learning. It predicts whether or not a star has an exoplanet orbiting it based on the change in flux (light frequency) data. The model used was a KNN and it obtained a 98% accuracy on the test set. The UI is a website that was made with Django, HTML, and CSS. It takes in a .csv file from the user and gets the change in flux data from that file. It then shows whether or not an exoplanet is orbiting the planet and a graph of the change in flux.
exopotato
App that trains itself using flux datasets which classify as having a planet near, or not. This allows for accurate estimation if planets are present near a star depending on a custom user-inputted flux dataset
FakeNewsDetection
One-Eye-is-All-You-Need-Lightweight-Ensembles-for-Gaze-Estimation-with-Single-Encoders
ReviewClassifier
ML model that classifies reviews as either positive or negative based on the data from the imdb dataset from keras.
StockTradingAI
This project is a Stock Trader trained to trade stocks from the S&P 500. It was made using a Deep Q-Learning model and libraries such as TensorFlow, Keras, and OpenAI Gym. It was trained on data from 2006-2016, cross validated on data from 2016-2018, and tested on data from 2018-2021
rishipython's Repositories
rishipython/StockTradingAI
This project is a Stock Trader trained to trade stocks from the S&P 500. It was made using a Deep Q-Learning model and libraries such as TensorFlow, Keras, and OpenAI Gym. It was trained on data from 2006-2016, cross validated on data from 2016-2018, and tested on data from 2018-2021
rishipython/One-Eye-is-All-You-Need-Lightweight-Ensembles-for-Gaze-Estimation-with-Single-Encoders
rishipython/BrainTumorSegmentationAI
This project uses the 2D U-Net architecture to make a machine learning model to segment a brain tumor in an MRI scan of brain. This project was done in Python and used various libraries and frameworks including Tensorflow (keras), numpy, pandas, and more. This program creates 4 different U-Net models, each with different depth (2, 3, 4, and 5). I then train these on the training set and evaluate them on the validation set.
rishipython/Acute-Lymphocytic-Leukemia-ALL-Cell-Classification
Due to morphological similarity at the microscopic level, making an accurate and time sensitive distinction between blood cells affected by Acute Lymphocytic Leukemia (ALL) and their healthy counterparts calls for the usage of machine learning architectures. However, three of the most common models, VGG, ResNet, and Inception, each come with their own set of flaws with room for improvement which demands the need for a superior model. ALLNet, the proposed hybrid convolutional neural network architecture, consists of a combination of the VGG, ResNet, and Inception models. The ALL Challenge dataset of ISBI 2019 contains 10,691 images of white blood cells which were used to train and test the models, 7,272 of cells with ALL and 3,419 of those that were healthy. Of the images, 60% was used to train the model, 20% for the cross validation set, and 20% for the test set. ALLNet outperformed the VGG, Resnet, and the Inception models across the board, achieving an accuracy of 92.6567%, a sensitivity of 95.5304%, a specificity of 85.9155%, an AUC score of 0.966347, and an F1 score of 0.94803 in the cross validation set. In the test set, the ALLNet achieved an accuracy of 92.0991%, a sensitivity of 96.5446%, a specificity of 82.8035%, an AUC score of 0.959972, and an F1 score of 0.942963. The utilization of ALLNet in the clinical workspace can better treat the thousands of people suffering with ALL across the world, many of whom are children.
rishipython/BostonHousingPricesPredictions
ML model that predicts Boston Housing prices based on data from the boston_housing dataset from keras.
rishipython/deepcgh3
DeepCGH3
rishipython/Exo-Potato
Exo-Potato is a software for deep space exoplanet hunting that uses machine learning. It predicts whether or not a star has an exoplanet orbiting it based on the change in flux (light frequency) data. The model used was a KNN and it obtained a 98% accuracy on the test set. The UI is a website that was made with Django, HTML, and CSS. It takes in a .csv file from the user and gets the change in flux data from that file. It then shows whether or not an exoplanet is orbiting the planet and a graph of the change in flux.
rishipython/exopotato
App that trains itself using flux datasets which classify as having a planet near, or not. This allows for accurate estimation if planets are present near a star depending on a custom user-inputted flux dataset
rishipython/FakeNewsDetection
rishipython/ReviewClassifier
ML model that classifies reviews as either positive or negative based on the data from the imdb dataset from keras.