In 2 hours 26 minutes, I am going to be 26 years old. I wanted to celebrate my birthday by taking a breath after holding it for 26 years.

Holding my breath, at the beginning

So, the ones who know me in person, know the story of my first name. Tekin is the name of the doctor who saved my life during my birth. I on’t want to give many details, but briefly, I did not breathe around 5 minutes when I was born. He made me breathe with artificial respiration. Then, my mom and dad gave me his name.

So, let’s start

So, after being a “good boy”…

A Beautiful Evening in Riga
Firstly, these white stars with yellow linings appeared. Their slow dance in three dimensional space was like a indication of miracle.

Then, I walked more and another street has come. I knew here from before, not that I passed this street before, but it was like a far memory hidden in my mind.
Deformed red lights hanging on the trees and street lights dance together on an arc.

Thylacine - Satie 1 plays, a sorrow starts in my mind. Thylacine continues with debussy, and tears appear on my eyes. But they disappear when I recognize…

What is the concept? How do we define it? How our consciousness perceive, record, and recall a concept? In this post, I will try to answer these questions.

Yesterday, I went out. I just wanted to go out in the fresh-cold streets of Riga. It is snow everywhere, therefore I decided to walk instead of biking. First, I walk down through Birznieka Upisa Street. When I reach the entrance of the bazaar, I turned right and got inside of it. When I reached Knockout barbershop, I turned left and got out of the bazaar from the parking area. I walked…

By using the system solution, it is possible to backpropagate the amount of error in convolution layers. In this post, I am going to talk about how to do it on the code in the Jupyter notebook.

Backpropagation methods in CNNs are commonly known methods. In my free time, I thought that there may be another way to backpropagate the error to the weights so that we can map the input to the desired output. Maybe there exists a method that gets all data and then makes some calculations and directly prints out the solution in one step instead of…

A small experiment to observe the effect of inequality on economic growth

I personally believe that economic inequality affects the growth of a country or the whole world negatively. Therefore, I thought that the intuition that I have should be proved by a small experiment. In this post, we will make the experiment together and observe the results.

Since we need to generate data realistically, let’s start defining our subjects first. I divided the citizens into three types which are low class, middle class, and high class. …

The value is only measured when you lose its source.

You are covered with people around you, your loved ones, and you never understand their value. You live in your country, in peace, with confidence and joy of communication, and you never know its value. You are grown with the thoughts and lessons of Ataturk, but you never understand its value.

And then, there comes a moment:

The moment that you are alone,

The moment that you are away in a foreign county.

Then you understand their value.

You lose everything, to gain them one more time. This time, by…

Links: Demonstration video & Github repository

Imagine that you are a baby that just opened your eyes to the world. Imagine the lack of consciousness, imagine not knowing anything. You don’t even know how to see something, let alone know how to recognize something. Now, come back to your consciousness. You can see, you can recognize any object around you and those objects remind you even of your memories.

But, how did you acquired all these skills? This is the starting point of Multi-Perspective Neural Networks(MPNN).

In my previous article, I have talked about an unsupervised learning algorithm that can…

Data is everything. Especially in deep learning, the amount of data, type of data, and quality of data are the most important factors. Sometimes the amount of labeled data that we have is not enough or the problem domain that we work on does not make sense to use a big amount of data such as in few-shot learning. The important factors in those situations are the algorithms. In deep metric learning, the loss functions that we use are the most important factors. …