Evaluation of GARCH Forecasting Performance Under Different Error Term Distributions

, M.Sc. Thesis Sharif University of Technology Khajian, Hamideh (Author) ; Zamani, Shiva (Supervisor)

Abstract

Volatility is the most important components in numerous finance applications. So, the methods of volatility forecast with reasonable accuracy require a deep attention.In this thesis with considering several distributions for error term, GARCH forecasting performance is evaluated on the intra- day data of "Foolad" stock returns by two loss functions of "MAE" and "HMAE". This evaluation is done in three forecast horizons, 1 day, 5 days and 20 days. Finally, the result of this study is as follows. GARCH (1, 1) forecast model with skewed t- student error distribution has the minimum value in the both loss functions for 1 day and 5 day forecast horizons. Also GARCH (1, 1) forecast model with t-...

Optimization of Water Consumption and Wastewater Treatment in Refinery

, M.Sc. Thesis Sharif University of Technology Khajian, Nafiseh (Author) ; Shayegan, Jalaloddin (Supervisor)

Abstract

Water is used as a key material in different processes of chemical industries. Considering abundant demand of chemical industries for water, pollution that threats this valuable material and the fact that there is currently no proper substitute for water, an urgent necessity is felt for economic utilization of water in industrial processes and optimal use of water resources. A variety of methods have been developed for analysis of water consumption in industrial processes, amongst are Pinch methodology and Mathematical Programming approach. In this thesis a Petroleum Refinery has been preferred as a case study in which optimization of water consumption in the aforementioned petroleum...

Generalization of the Online Prediction Problem Based on Expert Advice

, M.Sc. Thesis Sharif University of Technology Tavangarian, Fatemeh (Author) ; Foroughmand Araabi, Mohammad Hadi (Supervisor) ; Alishahi, Kasra (Co-Supervisor) ; Hosseinzadeh Sereshki, Hamideh (Co-Supervisor)

Abstract

One of the most important problems in online learning is a prediction with expert advice. In each step we make our prediction not only based on previous observation but also use expert information. In this thesis, we study the different well-known algorithms of expert advice and generalize problems when data arrival is in the two-dimensional grid. regret is a well-studied concept to evaluate online learning algorithm. online algorithm when data arrive consecutively in T time step has regret O (√(T)) . regret in two-dimensional grid with T row and P column is O(T√(P)).
2010 MSC: 68Q32 ; 68T05 ; 90C27