Large-scale Non-linear Regression within the MapReduce Framework By: Ahmed Khademzadeh Thesis Advisor: Philip Chan, Ph.D. Regression models have many applications in real world problems such as finance, epidemiology, environmental science, etc.. Big datasets are everywhere these days, and bigger datasets would help us to construct better models from the data. The issue with big datasets is that...

The paper compares several non-linear regression methods on synthetic data sets generated using standard benchmarks for continuous black-box optimization. For that comparison, we have chosen regression methods that have been used as surrogate models in such optimization: radial basis function networks, Gaussian processes, and random forests. Because the purpose of black-box optimization is freq...

Approximate Bayesian inference on the basis of summary statistics is wellsuited to complex problems for which the likelihood is either mathematically or computationally intractable. However the methods that use rejection suffer from the curse of dimensionality when the number of summary statistics is increased. Here we propose a machine-learning approach to the estimation of the posterior densi...

In this paper, we consider using a semiparametric regression approach to modelling non-linear autoregressive time series. Based on a ®nite series approximation to non-parametric components, an adaptive selection procedure for the number of summands in the series approximation is proposed. Meanwhile, a large sample study is detailed and a small sample simulation for the Mackey±Glass system is pr...

in the present study, the energetic and economic modeling of lentil and chickpea production in esfahan province of iran was conducted using adaptive neuro-fuzzy inference system (anfis) and linear regression. data were taken by interviewing and visiting of 140 lentil farms and 110 chickpea farms during 2014-2015 production period. the results showed that the yield and total energy consumption w...

In this work some quantitative structure activity relationship models were developed for prediction of three bioenvironmental parameters of 28 volatile organic compounds, which are used in assessing the behavior of pollutants in soil. These parameters are; half-life, non dimensional effective degradation rate constant and effective Péclet number in two type of soil. The most effective descripto...

Most works on Support Vector Regression (SVR) focus kernel or loss functions, with the corresponding support vectors obtained using a fixed-radius [Formula: see text]-tube, affording good predictive performance datasets. However, fixed radius limitation prevents adaptive selection of according to data distribution characteristics, compromising SVR-based methods. Therefore, this study proposes a...

A new kind of linear model with partially variant coefficients is proposed and a series of iterative algorithms are introduced and verified. The new generalized linear model includes the ordinary linear regression model as a special case. The iterative algorithms efficiently overcome some difficulties in computation with multidimensional inputs and incessantly appending parameters. An important...