Afshan Fathima

Highly skilled and result-driven Data Scientist with a passion for solving complex business challenges through data-driven insights. Proficient in applying advanced statistical analysis and machine learning techniques to extract valuable patterns and trends from diverse datasets. Possessing a strong programming background in Python and expertise in using PySpark, scikit-learn, Keras, TensorFlow for building machine learning solutions. Proficient at collaborating with cross-functional teams and effectively communicating to the stakeholders. Seeking to leverage my expertise to drive innovation and contribute to the success of data-driven organizations.

Architectures : K-NN , SVMs , RNN, NER , Sentiment Analysis , NLP 

Tech Stack   : Python, scikit-learn, Pandas , Keras , Tensor-Flow

MACHINE TRANSLATION 

Objective - Automated translation of source material into another language without human intervention.
 

- This project was designed to convert any input sentence from English to German or vice versa.

Techniques - RNN , LSTM , bi-directional LSTM , Encoder-Decoder

Tools used - Python , Keras , Scikit , TensorFlow 

HARD DRIVE FAILURE 

Objective - Tried to record and verify from the historical data as to which of the drives is likely to fail next.

-This project was based on a customer’s requirement as they had frequent drive failures in their storage array’s.

-This conclusion has helped them reduce the replacement time by 60% as most likely they would make an order for a new drive before the failure based on this analysis.

Techniques - Survival analysis - Kaplan-Meier method 

Tools Used - Python

MOVIE REVIEW ANALYSIS

Objective - To analyse each movie review and conclude if its a positive review or a negative review , also leading to the sentiment analysis of the writer and concluding on the type of the review.

Techniques - Sequential Model - LSTM

Tools Used - Python , NLP , RNN , LSTM