Dimitris Effrosynidis

Dimitris Effrosynidis

Mail LinkedIn Github Google Scholar Medium Kaggle

Experienced Data Scientist with a Ph.D. in Data Science, blending 6+ years of research expertise with 4+ years in business. Proven track record in analytics, programming, and modeling. Published 14 papers, executed successful projects, and showcased skills through blogging and GitHub. Dedicated to ongoing learning and staying abreast of industry trends.

Hard Skills

  • Data Mining
  • Data Analysis
  • Data Mining
  • Generative AI
  • Data Visualization
  • Machine Learning
  • Feature Engineering
  • Time Series Forecasting
  • Anomaly/Outlier Detection
  • Natural Language Processing
  • Supervised/Unsupervised ML

Tools

  • Python
  • Pandas, Numpy, SciPy
  • Scikit-Learn
  • SkTime
  • LightGBM, XGBoost
  • PyOD, Shap
  • Matplotlib, Seaborn
  • Plotly, Dash
  • Jupyter, Anaconda
  • Docker, Git
  • Django, React
  • MLflow, Airflow, AWS
  • MySQL, PostgreSQL
  • MongoDB
  • LaTeX, Overleaf
  • Notion, Jira, Confluence

Fine-Tuning a Pre-trained LLM for Sentiment Classification

less than 1 minute read

The project is available online on Towards AI.

In this tutorial, we will fine-tune the Task-Specific Sentiment Model (juliensimon/reviews-sentiment-analysis) and see if its 0.79 accuracy will improve.

Continue your read on Towards AI.

Leave a Comment

You May Also Enjoy

Mastering LLM Interactions

less than 1 minute read

Prompt engineering, quantized model optimization, and grammar-constrained sampling