Barbaros Şair's Projects
Predict which cell is cancerous with 96% accuracy using SVM machine learning algorithm.
Including ResNet50, AlexNet, LeNet-5, and VGG-16. This is written by me step by step, you can gain a better insight into these CNN architectures.
Predict carbondioxide emissions of cars based on their models, engine sizes, cylinders, etc. using regression. The model has a 88% accuracy. The dataset is present in the repo.
A convolutional neural network library written by me step-by-step using only numpy.
Predict if the customers will leave the service or not with 80% accuracy using Logistic Regression.
Detect and segment brain tumors precisely and fast.(accuracy=94%)
Pick the best drug for patients with 100% accuracy using decision tree algorithm.
This repository includes the building blocks of a general neural network. It is written by me step by step, and aimed to explain "the mysterious process" behind neural networks. You can find explanation of the algorithms like forward-propagation, back-propagation, gradient descent, and much more there! Take a look!
Predict house prices with over 75% accuracy using decision tree regression algorithm. The dataset is in repo.
This repository helps you to improve the performance of your own neural networks. You can learn about implementing the algorithms like L2 and dropout regularization, optimization algorithms like adam and momentum, mini-batches, parameter initialization, and more! Take a look if you want to train your model to be faster and more accurate!
This code intends to make the robot find its way in a maze. This code is an implementation of Tremaux Algorithm. This code guarantees to solve every maze available. You can find its predecassor in my previous repository.
A neural network library for Python. It is written by me step by step, and only uses the Numpy library.
Greetings from Istanbul, Turkey! I am 18 years old student interested in computational neuroscience, AI, neuromorphologic computing, and composing music. Currently, I study Electrical-Electronics Engineering in METU, the best technical university of Turkey according to QS, Times Higher Education, CWUR, et cetera. I am in Robotics Society of my university, and won various awards with my team, including in maze-solving and design-build categories. I am also Co-Founder of Parallel Technologies, which is a project team/enterprise based in METU. Moreover, I take courses from Coursera to advance my knowledge in this topics. Because to me, this topics hold the power of changing the world. You can find more about me in https://www.linkedin.com/in/barbaros-%C5%9Fair-5b373b21b. Please check it out!
Predict the services used by telecom customers based on their demographic data with K-Nearest Neighbors algorithm. The dataset is present in this repo.
This code is an application of Tremaux Algorithm. It was designed for a maze-solving robot attending to BROCUP competition. This code creates an array simulation to test the validity of the algorithm.
This code aims to follow the line between two lines. It outputs the degree between the middle line and the line which splits the frame in half vertically.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent