Artificial Intelligence and Innovation Centre (AIIC)

Tarik A. Rashid, PhD, PFHEA-UK

Professor in Computer Science/Artificial Intelligence – Director of Artificial Intelligence and Innovation Centre (AIIC)

Yassin Hussein Rassul

Research Assistant in Artificial Intelligence (AI)

Alan Salam Mohammed

Game Developer/Research Assistant

Hussein Mohammed Ali

Data Engineer / Research Assistant – Artificial Intelligence and Innovation Centre (AIIC)

Arwaa Wshyar Abdulkareem

Data Analyst/ Research Assistant – Artificial Intelligence & Innovation Centre

Muhammad Sabah Ibrahim

Data Engineer/ Research Assistant – Artificial Intelligence & Innovation Centre

Abdulhady Abas Abdulla

Artificial Liver Classifier: A New Alternative to Conventional Machine Learning Models

GitHub
The Artificial Liver Classifier (ALC), inspired by liver detoxification, offers simplicity, speed, and reduced overfitting. Trained with the improved FOX algorithm, it achieved high accuracy, including 100% on Iris and 99.12% on Breast Cancer, highlighting the promise of biologically inspired machine learning models.

Foxtsage vs. Adam: Revolution or Evolution in Optimization?

GitHub
Foxtsage, a hybrid optimizer combining FOX-TSA and SGD, improves neural network training by reducing loss mean by 42.03% and enhancing consistency over Adam. It also shows modest gains in accuracy, precision, recall, and F1-score but with a 330.87% increase in computational time, making it a robust alternative for optimization tasks.

Ratslayer

Ratslayer is a Vampire Survivors-inspired game where you battle waves of enemies as a quirky hero in a cartoony world, featuring unique power-ups and fast-paced action.

Zingil

Zingil is a language-based puzzle game that challenges players to solve linguistic riddles in an immersive world.

kurdishRoBERTa-language-detector-1B

This is a fine-tuned version of abdulhade/RoBERTa-large-SizeCorpus_1B, designed for detecting and classifying Kurdish and English text. Leveraging a custom bilingual corpus, this model is effective in distinguishing between these languages and accurately identifying text segments.

Kurdish-English Machine Translation with Transformers

This repository focuses on fine-tuning a Kurdish-English machine translation model using Hugging Face’s transformers library with MarianMT. The model is trained on a custom parallel corpus with a detailed pipeline that includes data preprocessing, bidirectional training, evaluation, and inference.

Kurdish Language Detector Model

This is a fine-tuned version of abdulhade/RoBERTa-large-SizeCorpus_1B, designed for detecting and classifying Kurdish and English text. Leveraging a custom bilingual corpus, this model is effective in distinguishing between these languages and accurately identifying text segments.

NER-FineTuing-RoBERTa

The project focuses on fine-tuning a pre-trained RoBERTa (Robustly Optimized BERT Pretraining Approach) model for Named Entity Recognition (NER) tasks. NER is a fundamental Natural Language Processing (NLP) task that involves identifying and classifying entities in text, such as names of people, organizations, locations, dates, and more.

NER

This dataset is designed for NLP tasks like sequence tagging and NER. It includes 1,472 sentences, 9,528 tokens, and 42 unique tags, making it ideal for training and evaluating NER models and token classification.

Corpus

Text corpus

5,721,505 number of rows

Kurdish large text corpus

A collection of Central Kurdish (Sorani) text corpus with a size of 430 MB

Number of rows: 773,800