Name: Izzet Turkalp Akbasli
Type: User
Company: Hacettepe University
Bio: Pediatric Resident, Artificial Intelligence, PICU and critical care enthusiast, JPIC Editor, Kaggle expert son husband, and father
Twitter: turkalpmd
Location: Ankara
Blog: turkalpmd.com
Izzet Turkalp Akbasli's Projects
I have completed my first project that machine learning on streaming data using Kafka and Docker. You can check-up my GitHub repository for codes.
AI-Driven Research Assistant: An advanced multi-agent system for automating complex research processes. Leveraging LangChain, OpenAI GPT, and LangGraph, this tool streamlines hypothesis generation, data analysis, visualization, and report writing. Perfect for researchers and data scientists seeking to enhance their workflow and productivity.
Automatically build ARIMA, SARIMAX, VAR, FB Prophet and XGBoost Models on Time Series data sets with a Single Line of Code. Now updated with Dask to handle millions of rows.
Building Data Science Applications with FastAPI, Published by Packt
As a result, it performed as well as other models. The worst disadvantage is the time! It takes almost four hours to run. Therefore, this algorithm still has a long way to go. However, it provides an alternative to standard algorithms.
CART analysis¶ As computing power and statistical insight has grown, increasingly complex and detailed regression techniques have emerged to analyze data. While this expanding set of techniques has proved beneficial in properly modeling certain data, it has also increased the burden on statistical practitioners in choosing appropriate techniques. Arguably an even heavier burden has been placed on non-statistician health practitioners – in university, government, and private sectors – where statistical software allows for immediate implementation of complex regression techniques without interpretation or guidance. In response to this growing complexity, a simple tree system, Classification and Regression Tree (CART) analysis, has become increasingly popular, and is particularly valuable in multidisciplinary fields.
Of all the applications of artificial intelligence, diagnosing any disease using a "black box" is always going to be a hard explanation. Those who will use the application will want to know how the model decides on the treatment conditions or following-up conditions according to the model result. Or data provider clinicians will want the model with the highest performance in their project. This dataset classified patients according to sacral position properties. I investigated using the below techniques for the best result and explainable machine learning model; Balancing unbalanced medical data Creating models with CatBoost Classifier Finding the most optimized parameters by Grid Search with the Optuna library Artificial intelligence algorithms described as Black Box are actually explainable SHAP library tutorial Combined use of RFECV and SHAP library for Feature Selection Comparison of all applied models to each other
the ChatGPT model, the AWS Lambda function, and the Streamlit interface with Docker.
Blog post on Medium
depremdoktor.site adresinde bulunan ilaç hesaplama ve demprem ilişkili kliniklerin yönetimini içerir
This code is working effeciently on pubmed for PubMed display option.
Our end-to-end pipeline is designed to flexibly process data from different recording machinery and to read data in PDF format as well as data from native digital devices delivered in XML.
I am really curious that, how I must create categorical features from numeric features. The most commonly used method is separating with the same intervals and stratification with quantiles. But my experience in medicine showed that this stratification threshold is wrongly chosen. For example, I have already dropped some values in the "BMI" feature that are bigger than 60 and smaller than 14. But some notebooks include them and are replaced them with some values. But other hand, when I used medical categorization guides also doesn't result in good model performance.
A HYBRID APPROACH TO ANOMALY DETECTION USING FUZZY LOGIC TUNED WITH EVOLUTIONARY ALGORITHMS
Build large language model (LLM) apps with Python, ChatGPT and other models. This is the companion repository for the book on generative AI with LangChain.
In this repository, I will present a tutorial of Genetic Algorithm.
Jupyter Notebook tutorials on solving real-world problems with Machine Learning & Deep Learning using PyTorch. Topics: Face detection with Detectron 2, Time Series anomaly detection with LSTM Autoencoders, Object Detection with YOLO v5, Build your first Neural Network, Time Series forecasting for Coronavirus daily cases, Sentiment Analysis with BER
My primary goal in this notebook, is we have a lot of algorithms for solutions, just know what you are searching for
I am trying to forecast COVID disease progress with time series algorithms
My first web-page project
Parallel Hyperparameter Tuning in Python