Matt-Jennings-GitHub/Skin-Cancer-Identification Matt-Jennings-GitHub/Skin-Cancer-Identification

  • Stars
    star
    3
  • Rank 3,944,587 (Top 79 %)
  • Language
    Python
  • Created about 5 years ago
  • Updated about 5 years ago
Twitter logo Share LinkedIn logo Share Facebook logo Share
Matt-Jennings-GitHub/Skin-Cancer-Identification
github

Matt-Jennings-GitHub

Reviews

There are no reviews yet. Be the first to send feedback to the community and the maintainers!

Repository Details

Image recognition using a CNN (both with and without the aid of transfer learning) to detect the presence of melanoma in images of malignant and benign moles taken from the ISIC Medical data archive. (Keras, TF backend)

More Repositories

1

Fishing-Game-NEAT

Teaching an AI to play a recreated and simplified version of the popular ice fishing game from Club Penguin using a neural network trained with genetic evolution using the NEAT algorithm.
Python
9
star
2

ConnectFour-ComputerVisionAI

A computer vision project to detect the presence of the board game "ConnectFour" ("Four-In-A-Row") from an image, and identify the next player's optimal move. The image is processed with the Python implementation of CV2 and we perform Bilateral Filtering, Edge Detection, Contour Analysis and HSV Filtering to identify the current game state. Next, we apply the MiniMax algorithm with alpha-beta pruning to determine the best move for the next player to make!
Python
4
star
3

Tata-Global-Forcasting

Forecasting closing stock prices from a dataset of past 'Tata Global' stock data obtained from Quandl. Applying various ML algorithms including Linear Regression, KNN, ARIMA and, most successfully, a LSTM deep neural network. (Keras, TF backend) achieving an impressively low rmse.
Python
2
star
4

Cocktail-Classifier

Image recognition with CNN trained on google images to identify types of cocktails from images. Network written in Keras (TF Backend) and applied transfer learning from VGG-16 trained on the imagenet dataset.
Python
2
star
5

Coronavirus-Data-Visualization

Interactive live data visualisation project to show the global spread of the Coronavirus. Includes geographic mapping of the COVID-19 epidemic, representing country cases and deaths using a colour heat matrix. Overlay of economic data on closing price of major world indices to show the economic impact of the crisis. Data sourced from the ECDC and Yahoo Finance.
Python
2
star
6

Matt-Jennings.github.io

Project Portfolio Website
JavaScript
1
star
7

Digit-Classification

Applying a CNN written in Keras (TF backend) to classify digits from the MNIST dataset.
Python
1
star
8

Movie-Recommendation-System

System to recommend movies to users based on dataset of past movie attributes and ratings using low rank matrix factorisation.
Python
1
star
9

House-Price-Prediction

Applying Gradient Boosting Regression to predict the predict the value of housing from dataset of past house sales and their attributes. Using Keras (TF Backend) Including preprocessing features, visualizing the dataset and running the model on GCP.
Python
1
star
10

Coronavirus-Prediction

End-to-end machine learning project. Predicting and understanding the spread of the COVID-19 Pandemic in the USA can help forecast supply chains and save jobs and lives. The past reproduction number is first calculated using estimates based on the SIRf pandemic model. Geographic, Socioeconomic and Health factors from individual counties are sourced and analysed from USA government datasets. These factors are used in machine learning models to predict the spread of the virus. The final results are visualised geographically.
Jupyter Notebook
1
star