Ca'Foscari of Venice
In recent years an active area of research emerged in economics and finance, which aims at handling inference on high dimensional and complex data by combining machine learning methods with econometric models. This project aims to investigate the application of dimensionality reduction methods in forecasting with large panel and builds on the literature on traditional (such as Principal Component Analysis, Independent Component Analysis) and relatively new dimensionality reduction methods (such as Sammon projection, Multidimensional Scaling, Isomap, Laplacian eigenmaps, and Local linear embedding, Random Projection, Kernel PCA, Nonlinear Factor Analysis), available from machine learning and statistics. Under a methodological point of view, given to the complexity of the challenges, this project will focus on some of these dimensionality reduction techniques, with the aim of proposing new econometric models for large dimensional data and assessing the properties of the estimators and predictors for the proposed new models This project aims to investigate the application of dimensionality reduction methods.
A public selection is called for the award of n. 1 allowance for 12 months for post-doc, researchers and professors. The research program foresees the beginning of the activity for the month of October 2020.The amount of the cheque is Euro 9.541,00 per year (net of expenses load of the paying agency).
29/9/2020 – 20/10/2020