Posts

Why ANN does not perform well on sequential data?

Because simple feed forward network assumes that data points are independent and treats each data point individually while sequential data is characterized by dependency, that is each data point is some how dependent on previous data point. Hence feed forward network fails to learn dependency information and consequently results in poor predictions. Here comes RNN for rescue yay.

Why Tensorflow?

Because it provides  extensive built-in support for deep learning mathematical functions for neural networks auto differentiation (for back propagation) and optimizers Source

Different ways to create pandas dataframe

Image
Image by Gerd Altmann from Pixabay Click here to view notebook file

How to normalize dataframe column(s) values in range 0-1

How to resolve anaconda-navigator: command not found error

If anaconda navigator is freshly installed on system then run this in terminal $ export PATH=/home/userName/anaconda3/bin:$PATH else run below $ source ~/anaconda3/bin/activate root $ anaconda-navigator

Converting date time timestamps to Unix time

Image
Photo by Nathan Dumlao on Unsplash Here we will see how to convert date time timestamps to Unix time. Unix stores time as number of seconds that have elapsed since the Unix epoch. Unix epoch is the time 00:00:00 UTC on 1/1/1970. We first generate timestamps using: Now we convert these timestamps into Unix time using:

Basic introduction to working with time series data

Image
Image by annca from Pixabay This is a basic introduction to working with Time Series data using Pandas which is a powerful python library for data manipulation. Let's start with creating a data frame. For the sake of simplicity, our data frame consists of only two columns.   Our data set range can be seen here. If we want to view data from 11th to 20th minute, we can do this using: To view data of first 59 minutes only: We can do above for year or month as well.