House Price Prediction

25 Aug 2018

Introduction

The problem involves analyzing housing prices that were sold between May 2014 and May 2015. In this post, we are going to perform two task. 1) exploratory data analysis 2) use machine learning model to predict the sale price fo the houses.

Air Pollution in France

20 Aug 2018

Time Series

15 Aug 2018

what is stationary and why do we need it?

Strongly stationary Definition: $F_X(x_{t_1+\tau},...,x_{t_k+\tau})=F_X(x_{t_1},...,x_{t_k}) \space for \space all \space k \space ,\tau$ This means the joint distribution of any length of the sequences is the same if we shift the window by any time gap.

Pits of image segmentation

06 Aug 2018

Tree Detection Project

Memory Error

The satellite images are too big to copy to GPU or to perform convolutions on. A typical type of an image is 100K * 100K * 4 pixel. One natural solution for this is to divide the image in a predefined number of tiles and train on small batches.