Machine Learning Process design and implementation in Python

Hi, The project is to build a Machine Learning algorithm applied to financial time series developed in Python composed with the followings (this algorithm may be hybrid, using different methods): - Financial time series adjustments module in Python. (i.e Splits and Dividends adjustments on stocks time series, Future roll-over adjustments module using different methods...) - Meaningful feature selection and importance assessment module in Python (There will be several time series to predict, all using the same global datasets to train, validate and run) - Hybrid or composed Machine Learning Process to reach targeted output with a defined tolerance to error. - Performance or Trust/confidence measure of the output. Then, once the algorithm is ready, organise and make the necessary arrangements complying with defined constraints to set it up to production on a daily basis. The development/research datasets will be provided through csv files containing the input time series. A high level of discretion is expected and NDA will be required. Looking forward to considering your application Sylvain Director