Unlocking the Power of AI and ML: My Tale of Art and Science with Dlithe
Hi there! My name is Ananya Pai, and I recently completed an AI and ML internship. As a second-year student pursuing a degree in Computer Science , I’m excited to share my experience with you.
The one-month internship program was an incredible opportunity for me to gain practical experience in the field of Artificial Intelligence and Machine Learning from DLithe. Throughout the internship, I worked on a variety of projects and tasks that helped me build my skills and develop a deeper understanding of AI and ML concepts.
Summary
During my AI and ML internship with Dlithe, I first learned Python, which is the most popular language for AI and ML. Python is widely used in data science, machine learning, and artificial intelligence because of its simplicity, readability, and vast library support.
Once I had a good understanding of Python, I was introduced to the basic concepts of AI and ML, including supervised and unsupervised learning, classification, regression, clustering, and more. I learned about different machine learning algorithms and how to select the best algorithm for a particular problem. I also learned about data preprocessing techniques, such as data cleaning, feature scaling, and handling missing values.
During the 15-day online training, I gained an understanding of how to use various Python libraries, including NumPy, Pandas, Matplotlib, and Scikit-learn, which are essential for AI and ML. NumPy is used for numerical computations, Pandas for data manipulation and analysis, Matplotlib for data visualization, and Scikit-learn for implementing machine learning algorithms. Later, I was given three projects as self-assignments. These projects were designed to help me apply the concepts I had learned during the 15-day online training and build my skills in the field of AI and ML.
Projects Undertaken
The first project was to predict whether breast cancer was benign or malignant using logistic regression. This involved comparing the results with KNN and Naive Bayes classifiers, which I also implemented. The dataset used for this project was obtained from the UCI Machine Learning Repository. To complete this project, I had to apply my knowledge of supervised learning, logistic regression, KNN and Naive Bayes classifiers. I was able to gain a deeper understanding of these concepts through this project and learned how to effectively compare different machine learning models.
The second project was focused on text analytics, specifically web scraping a news article and performing sentiment analysis. Using BeautifulSoup, I gathered information from a website and used the text data for sentiment analysis. I made use of the Natural Language Toolkit (NLTK) and any classifier to analyze the polarity of the data (Positive, Negative, or Neutral). The resource used for this project was MonkeyLearn. Through this project, I was able to learn about web scraping, data preprocessing, and text classification. I also gained valuable insights into how sentiment analysis can be applied to different domains.
The third project involved clustering credit card users using a dataset obtained from Kaggle. To complete this project, I had to apply my knowledge of unsupervised learning and clustering algorithms. I used K-means clustering to group similar credit card users together based on their spending patterns. Through this project, I was able to understand the importance of unsupervised learning and clustering algorithms in data analysis.
Results
For the first project, we used logistic regression, KNN, and Naive Bayes classifiers to predict whether the cancer is benign or malignant. The accuracy of each classifier was evaluated, and the results were compared. The logistic regression classifier achieved the highest accuracy of 68.89%, followed by KNN with 67.50%, and Naive Bayes with 56.72%.
In the second project, we conducted sentiment analysis on a news article using NLTK and a classifier. The polarity of the data was analyzed and classified as positive, negative, or neutral. The sentiment analysis revealed that the text was mostly neutral, with a slightly positive sentiment. The compound score of 0.9993 suggests a very strong positive sentiment overall.
In the third project, we clustered credit card users using K-Means clustering algorithm. We analyzed the data and identified patterns and clusters of users based on their spending behavior. The clusters were visualized using a scatter plot, which helped us identify potential target groups for marketing campaigns.
Overall, these projects were instrumental in helping me apply the concepts I had learned during the AI and ML internship with Dlithe. They helped me gain practical experience and build my skills in AI and ML. Through these projects, I was able to unlock the power of AI and ML and gained valuable insights into how they can be used to solve real-world problems.
Conclusion
In conclusion, the AI and ML internship was a great opportunity for me to gain hands-on experience in this field. Through this internship, I learned the basics of Python programming and important concepts of AI and ML, including regression, classification, clustering, and sentiment analysis. The experience I gained during this internship will be immensely useful for my future academic and professional endeavors. The skills and knowledge I acquired will enable me to tackle real-world problems using AI and ML techniques. The training and projects were well-structured and provided a great learning experience. I believe that this internship will be worth trying for anyone who is interested in exploring the world of AI and ML.
Checkout my projects @ github.com .
