Presented By: Dr Malvina Marchese, Cass Business School, City, University of London
In the last decade, Machine Learning models have established themselves as serious contenders in the area of forecasting to classical econometric models. The aim of this advanced machine learning course with EViews is to offer an overview of the most popular Machine lLearning methods for prediction including neural network, decision trees, K nearest neighbourhood, kernel regressions.
The course offers plenty of practical examples that apply to financial Big Data and discusses comparison with traditional econometric forecasting methods such as ARIMA, Markov switching, smooth transition autoregressive models.
The course is EViews based and makes large use of the EViews-Python integration. Participants are not required to have any previous knowledge of python, but should have installed it before joining. The course is intended as an advanced course and attendance to the introductory course, machine learning in EViews, is highly recommended.
Participants will leave with an in-depth understanding of forecasting with machine learning , with the ability to run any python code via the user friendly and graphically advanced EViews interface , with basic python coding skills and plenty of codes examples.
Neural Networks: starting values, over fitting, input scaling, bagging.
Neural network : Bayesian neural networks and performance comparison.
Random forests: definition, out of bag sample, variable importance , overfitting , variance and de-correlation effects.
Knowledge of time series econometrics is desirable, although not required. Attendance to the introductory course Machin learning in EViews is strongly recommended.
No pre-reading is needed; references for post-course reading will be provided during the course itself.
Participants must have installed Python before joining the course. Previous knowledge of EViews is required.
The number of attendees is restricted. Please register early to guarantee your place.