I would like a programme built to forecast crypto currency prices using time series forecasting and machine learning.
The programme will involve: 1) importing cryptocurrency data (5000-11000 types) 2) capability to import other time series data like volume traded on a specific day to train models with 3) training data set, and predicting using test data 4) tuning model algorithms 5) implementation using multiple different machine learning methods: TBATS, LSTM, CNN, XGBoost, MLP, RBF, KNN, CART, GP, RNN, FFNN, ranger, ARIMA, SARIMA, VAR, HWES, seasonal decomposition 6) predict crypto currency price for next few time steps 7) repeat step 3-6 for multiple crypto currency assets, and rank assets according to highest accuracy of prediction
A detailed description of programme requirements will be sent after you accept the project.
Please answer the following questions to apply for the project: 1) Do you have the knowledge and capability to implement the programme using all the machine learning methods mentioned? 2) Can you demonstrate your knowledge or capability through a script or programme you have previously made? 3) How long will you require to complete the project?