Tech Tip's : Do's and dont's


February 2016

Machine learning Tips for Audio, Image and Video Analysis

Original post :-

By Lillian Peirson

Post is as follows : –

Neural networks are great in image, video, and audio machine learning problems. For example, if you have an image classification task, you can use convolutional neural nets. First, you’ll need to normalize your image, and then downsample it to a smaller size. Usually 16 – 64 pixels for each dimension is good.
After that you can build a simple convolutional net to learn from these downsampled images. The most important hyperparameter is the learning rate – tune it first. After that you can play around with changing layer sizes, the convolutional layer kernel, and pooling sizes. Try adding more layers and activation functions. Definitely try using the dropout method.
If your dataset is not very large, use data augmentation. Usually if you rotate your image or move it by a few pixels horizontally or vertically, the class doesn’t change, right? Sometimes you can even make a mirror image! Data augmentation can help you avoid some overfitting, making it possible to try an even bigger net. Finally, if you need a little better quality, you should definitely try to build several models with similar hyperparameters and then build a voting classifier on top of them.


Tech Tips: Do’s and Dont’s featured on

Hello readers,

I am excited to bring you this news that Tech Tips: Do’s and Dont’s is now featured on

Andrei, a moderator at Tech 3dge, presented me with an opportunity to feature this blog on their website.

Tech3dge is a website that features the newest electronic product updates and innovations, and EEWeb’s Europe website.

Their primary website is Elektronik Forum EEWeb.

EEWEb Europe is an electronics and electrical engineering online community with electronics forum and electrical engineering forum. It has electrical and electronic design articles, and resources
in the area of RF design, analog design, embedded design, PCB design, test and measure.

Link featuring this blog is present here

Thanks Andrei for featuring this blog.

To find more latest tech updates on Electronics and Electrical engineering community head over to their primary website and

Machine learning : SVM

SVM – Support Vector machines.

Different kernels – linear, rbf, poly etc.

Each Kernel is an additional parameter to SVM classifier to map data points to ta new dimensional space and then perform the classification.

Gamma more = ?

C value is more means that the error is more => More data points are correctly classified.

More the value of C does not gurantee an accurate classifier.

One needs to be careful of overfitting in machine learning.

For SVM overfitting the parameters in control are C, gamma, kernel used.

Indentation of files in vim.

We all wish to indent files in vim so that the code looks more uniform and coherent across multiple systems.

Here’s a script written to indent file which runs on bash. This script expects an input (argument) from the user which can be file name or a directory path [assumed that user wishes to run the script for all files in that directory]

#Pass absolute file name which you wish to indent as an argument to this script
if [[ -d $argument ]]; then
    echo “$argument is a directory”
    for file in `find $argument -type f -name “*.c” && find $argument -type f -name “*.h”`
        echo $file 
        vim $file -c “normal gg=G” -c “wq”
        expand -t 4 $file $file      
elif [[ -f $argument ]]; then
    echo “$argument is a file”
    vim $file -c “normal gg=G” -c “wq”
    expand -t 4 $file $file
    echo “$argument is not valid”
    exit 1

Save the above code as

Type ./ .

Blog at

Up ↑