Implementation of new strategies for the OLPS toolbox
New types of strategies for on-line portfolio selection in computational finance.
Our developers have developed new types of strategies for on-line portfolio selection in computational finance.
The first one is "Follow the leading history" (FLH) method described in "Efficient learning algorithms for changing environments" article by E. Hazan and C. Seshadhri. Core of the strategy lies in selection of an optimal solution among several ones suggested by several expert algorithms. Each of them uses the same optimization method but every expert makes calculations based on data from different periods. Selection of the optimal number and type of experts is specified in description of the strategy.
The second implemented strategy is Commission Avoidant Portfolio Ensembles (CAPE) described in "Online Learning of Commission Avoidant Portfolio Ensembles" article by G. Uziel and R. El-Yaniv. In this strategy along with existing expert algorithms there's one more artificial expert added. This artificial expert allows to keep existing portfolio unchanged and to avoid expenses for making operations, maximizing profit in a long-term outlook.
Both strategies are developed in Matlab. We've also built wrappers that allow to use them in OLPS toolbox (http://olps.stevenhoi.org/).
Model for restoring blurred and pixelated faces in a photo.
Workout helper app
Mobile app for the estimation of proper body positions during the workout.
Car rental price simulation and prediction
The goal of the project is to train models for car rental price prediction in Japan based on the prices and demand for car rental from some Japanese car rental companies and the history of weather data in Japan.