All my work related to data science, machine learning, deep learning and similar.
You can find a summary of each of the projects in their folders.
- Spam detection models evaluation
The task was to evaluate and measure different classification models in task of detecting spam e-mails based on data from SpamAssasin. The task was to tune classifiers in order to achieve desired recall and precision instead of accuracy. The data was also a little imbalance and many different approaches were used to conduct the sensitivity studies.
- Toxic text classification visualisations
Big data visualisations - multiclassification of 6 types of toxicity: ['toxic', 'severe_toxic', 'obscene', 'threat', 'insult', 'identity_hate'].
- K-NN evaluation and benchmarks
Benchmark K-NN classifier in supporting identification of myocardial infarction.
- ARQ protocol analysis in Matlab
Benchmark of ARQ protocol used in data correction during transmission.
- ETL and OLAP multidimensional data analysis
Multidimensional analysis with SSIS and SSAS of dean's office data.