Jakob Balkovec

As a dedicated computer science student-athlete, I actively engage in effective communication with teammates and coaches, utilizing strong diplomatic skills to collaboratively solve challenges. I efficiently balance a rigorous class schedule with athletic commitments, showcasing determination and effective time management. My self-motivation drives me to consistently work towards achieving ambitious goals. Known for my keen perception and attention to detail, I apply analytical thinking to comprehend work outcomes and strategize effective pathways for success. Striving to enhance my skills daily, I am committed to continuous improvement, embodying the qualities of a proactive and driven individual.

• Deutsche Sprach Diplom, Gimnazija Franceta Preserna, Kranj. Acquired in 2021

IMDb ML Model

This project showcases my expertise in machine learning, specifically in sentiment analysis applied to movie reviews from the IMDb dataset. The model utilizes a basic neural network architecture, providing insights into sentiments expressed in movie reviews.

Demonstrates my ability to construct, train, and evaluate a sentiment analysis model using the IMDb dataset. The model, a basic neural network, comprises an embedding layer, flattening layer, dense layer, and a sigmoid output layer.

link: https://github.com/j-balkovec/CodeHub/tree/main/imdb_model

Hidden Markov Model

Developed a Hidden Markov Model (HMM) class, a sophisticated tool tailored to model intricate systems characterized by hidden states and observable outputs. This HMM class stands as a testament to my prowess in probabilistic modeling and machine learning.

Equipped with an array of functionalities, the HMM class serves as a versatile solution. Its methods include the computation of forward and backward probabilities for observation sequences, enabling a comprehensive understanding of the system's dynamics. Moreover, the class incorporates the expectation-maximization algorithm, facilitating the automatic learning of model parameters based on observed data.

One of the standout features of this model is its predictive capabilities. By leveraging the learned parameters, the HMM class excels in predicting the most likely sequence of hidden states when provided with an observation sequence. This predictive power empowers the model to make informed decisions and uncover the underlying dynamics of complex systems.

link: https://github.com/j-balkovec/CodeHub/tree/main/hidden_markov_model

Natural Language Processing Chat Bot

This project showcases my expertise in sentiment analysis, featuring an algorithm for determining sentiment polarity in textual data. Notably, it integrates a visually appealing GUI crafted using the customtkinter module. The algorithm goes beyond basic classifications, offering nuanced insights into emotional tones with proficiency in identifying neutral expressions. The GUI reflects my commitment to delivering both functional and visually engaging solutions, enhancing the user experience. Its versatility is evident in handling diverse textual data sources, making it adaptable for various scenarios. The modular design allows seamless integration into different projects, emphasizing ease of use. Prioritizing functionality and user experience, the script provides an intuitive interface for users of varying technical expertise to effortlessly navigate sentiment analysis results. This project exemplifies my dedication to delivering high-quality, multifaceted solutions, contributing to the intersection of technology and user experience.

link: https://github.com/j-balkovec/CodeHub/tree/main/NLP_bot

Sentiment Analysis Algorithm

Harnessing the capabilities of Python, I've meticulously crafted a sentiment analysis algorithm that leverages the Natural Language Toolkit (NLTK) library, showcasing my adeptness in both programming and natural language processing. This algorithm serves as a powerful tool for classifying textual data based on sentiment polarity, contributing to the realm of data analysis and interpretation.

Operating seamlessly with data stored in a .csv file, the algorithm employs NLTK's robust features to delve into the intricacies of language. It performs sentiment analysis on the textual content, employing advanced techniques to classify sentiments as positive, negative, or neutral. This process not only demonstrates my coding proficiency but also underscores my commitment to extracting meaningful insights from diverse data sources. The Python-based sentiment analysis algorithm stands as a testament to my ability to integrate programming skills with natural language processing, enhancing my toolkit for comprehensive data analysis.

link: https://github.com/j-balkovec/CodeHub/tree/main/Sentiment_Analysis_Algorithm