Murat A. Avcu

I just published Python NLP : Duygu Analizi (Sentiment Analysis) 

7

"Retrieval-Augmented Generation-based Relation Extraction" Sefika Efeoglu, Adrian Paschke New progress in Retrieved-Augmented Generation for LLMs! 🤖 Information Extraction (IE) is the process through which unstructured text data is converted into structured data by employing entity and relation extraction (RE) methodologies. Large Language Models (LLMs) have emerged as promising solutions to the limitations of labeled data and RE accuracy, although they still face the problem of being prone to response hallucination. To overcome this, researchers at the Technical University of Berlin and Free University of Berlin propose Retrieved-Augmented Generation-based Relation Extraction (RAG4RE) approach, a new promising solution to the difficulties currently faced by traditional RE techniques. Consisting of an information retrieval, data augmentation, and generation modules, this RAG-based RE approach is able to outperform Vanilla LLM prompting and even other existing RE approaches, particularly evident in its performance on benchmark datasets like TACRED and TACREV. Moreover, it also mitigates the issue of hallucination. By enriching the context of LLM prompts through RAG, this approach enhances the accuracy and relevance of responses, paving the way for more efficient and effective information extraction. This research marks a significant step forward in the field, offering insights into the potential of RAG-based approaches to redefine how we tackle RE tasks in the era of Large Language Models. Therefore, this sets a new standard in natural language processing and new opportunities in the development of RAG-LLMs. Preprint: https://lnkd.in/gWhvyqrE #LLM #RAG #InformationExtraction #ResearchPaper

1

Dear connections, good evening to all, I want to share a project I have been working on recently with you. This project is based on a dataset that examines the purchasing intentions of customers for electronic products. Before starting to create the model, I processed the data in detail to make it suitable for machine learning models. After testing two different models (KNeighborsClassifier and RandomForestClassifier), I chose the RandomForestClassifier algorithm, which achieved a very high success rate of 95%. You can access the project codes and more details through the following Github and Kaggle links. https://lnkd.in/dMASccSv https://lnkd.in/d8-hkywa #FeatureEngineering #EDA #ExploratoryDataAnalysis #NewFeatureForDatasets #MachineLearning #HighAccuracy #DataAnalysis #DataScience #DataAnalyst #DataScientist

5

Just finished the course “What Is Generative AI?” by Pinar Seyhan Demirdag! Check it out: https://lnkd.in/g8cXg8nA #generativeai #artificialintelligence #generativeaitools.

14

2 Yorum

Matematiksel saat - - - - --- If you’re tired of paywalls cutting off your reading experience, try ProReader.io ------- 📌 Follow Abdul Rahim Roni for more content like this: If you like this and want more AI tools, tutorials and news, join Cool Tech Gadgets - my newsletter that helps you be smarter about Microsoft and AI and Google and technology: 👉 https://lnkd.in/grn4UX-8 #Engineering #Innovation #technology #engineer #Automotive

4

Hello everyone, Meet my new project, the house price prediction. In this project I did detailed feature engineering, exploratory data analysis and machine learning modeling. I used the Gradient Boosting Regressor, which has the highest accuracy score among the seven algorithms. You can access the details and codes of the project via the link below. Thanks. #FeatureEngineering #ExplorataryDataAnalysis #MachineLearningProjects #HousePricePrediction https://lnkd.in/dDg7md4T

12

Just finished the course “What Is Generative AI?” by Pinar Seyhan Demirdag! Check it out: https://lnkd.in/gPVwXJB8 #generativeai #artificialintelligenceforbusiness #artificialintelligence.

5

🚀 Unlocking the Future of Machine Learning with Natural Learning (NL) by Hadi Fanaee, Associate Professor of ML🚀 NL, inspired by prototype theory, is shaking up the machine learning landscape! Say goodbye to complex black-box models and hello to clear, interpretable insights. With NL's intuitive training algorithm, understanding AI has never been easier or more accurate. Key Takeaways: 🔍 Prototype Power: NL draws inspiration from prototype theory, offering a fresh perspective on machine learning. 💡 Simple Yet Mighty: NL's straightforward approach delivers powerful models with crystal-clear explanations. 🌟 Outshine the Rest: NL outperforms black-box models and traditional methods, setting a new standard for transparency and performance. 🌎 Versatile Applications: From healthcare to finance, NL is making waves in diverse fields, offering invaluable insights and fair explanations. 👩‍💻 Contributing to AI: NL's contribution lies in making AI more human-friendly and understandable, propelling us towards a brighter future in machine learning. Natural Learning: https://lnkd.in/gQt_vyjS Let's embark on the NL revolution and decode the mysteries of AI together! #NaturalLearning #MachineLearning #DecodeAI #Innovation #AIRevolution 🌐🔍🚀

9

part 2 recipe recommendations system using machine learning #machinelearning #recommendations #python #nlp #deeplearning #ai https://lnkd.in/dbQYyArA

2

I really enjoyed presenting my workshop on graph databases and Cypher vs. SQL at TMLS in Toronto a couple weeks ago. To all those who attended the session in person, many thanks for your feedback during/after, and for your requests to get a re-watchable recording! To all those who couldn't attend the session, I've just released the next video on our "Getting started with Kùzu" series on the Kùzu Inc. YouTube channel that covers part of what I demoed. The main focus of this video is to answer the question: "When do you need graph modeling & querying". The aim is *not* to state that Cypher is better than SQL, but rather, to highlight the scenarios when Cypher (and graph databases) are a better fit for the query workloads you may be interested in running. I show a comparison of using Cypher and SQL to answer questions about a financial network using a combination of command line queries and interactive queries that visualize the data as a graph. Along the way, some interesting insights are revealed in the data, which are enabled by the power of graph modeling. Hope this video helps those who are getting started with graph databases and Cypher! Do reach out with any questions or if something wasn't clear - always happy to hear feedback, so post your comments below 👇🏽. pip install kuzu Video link: https://lnkd.in/gmit4guB #graph #database

29

1 Yorum

Just finished the course “What Is Generative AI?” by Pinar Seyhan Demirdag!  This course explored the fundamentals of Gen AI leading into concepts such as Generative Adversarial Networks (GAN's), Variable Auto Encoders (VAE) for anomaly detection. It also touched upon the various Gen AI tools like Stable Diffusion, Midjourney and DALL-E. This was exciting and highly informative at the same time. Remember, the power behind AI is you. Check it out: https://lnkd.in/gHRWJ7Bv #generativeai #artificialintelligence #generativeaitools.

15

This latest video in Semih Salihoğlu graph database fundamentals series gives a truly excellent intro to RDF! It gently introduces RDF terminology along with concrete examples to help understand the benefits of #rdf and where it can be useful. Some Kùzu Inc. resources are also linked in the video description. Having studied these topics for a while myself, I remember how intimidated I was in my early days because it all felt so overwhelming, so I think this video will be very helpful for those who are getting started on their journey understanding the trade-offs between property graphs and RDF :). Looking forward to sharing part 2 of this video soon! #graph #database

5

🚀 Project Completed! 🚀 I'm thrilled to announce that I've successfully fine-tuned the FLAN-T5 Small model using QLoRA for question-answering tasks, working over weekends! 🔍 Overview: Data Collection: Scraped and curated data from websites. Fine-Tuning: Enhanced the FLAN-T5 Small model with QLoRA and generated question-answer pairs. Evaluation: Achieved significant improvements in model performance. Check out the code on GitHub: https://lnkd.in/dr-6weVF This project sharpened my skills in data scraping, NLP, and machine learning. Excited to apply these insights in future endeavors! #NLP#LLM #ProjectCompletion #GitHub #FLANT5 #QLoRA

19

Microsoft is training its own large language model, internally labaled MAI-1, with Mustafa Suleyman leading the effort. The model is around 500B parameters and could compete directly with LLMs from Google, OpenAI, etc. This marks the first time Microsoft has developed an in-house AI model of this magnitude since investing over $10 billion in OpenAI for the rights to reuse the startup's AI models. OpenAI's GPT-4 powers not only ChatGPT but also Microsoft Copilot. The development of MAI-1 is being led by Mustafa Suleyman, the former Google AI leader who recently served as CEO of the AI startup Inflection before Microsoft acquired the majority of the startup's staff and intellectual property for $650 million in March. Although MAI-1 may build on techniques brought over by former Inflection staff, it is reportedly an entirely new large language model (LLM) With approximately 500 billion parameters, MAI-1 will be significantly larger than Microsoft's previous open source models (such as Phi-3).

48

3 Yorum

🚀 Breaking Down the Attention Mechanism in Neural Machine Translation 🚀 In 2015, Bahdanau et al. introduced an incredible innovation in the world of Neural Machine Translation: the Attention mechanism. 🧠✨ It's a cornerstone referenced in the "Attention is All You Need" (Vaswani et al., 2017) paper that pioneered the Transformer model! #AttentionMechanism #Seq2Seq #MachineTranslation #NeuralNetworks #DeepLearning #Transformers #Alignment #NLP

11