A personal reflection on the benefits and challenges of taking an online course on data science and machine learning

Introduction

I have always been fascinated by artificial intelligence (AI) and its applications in various domains. However, my first attempt to learn AI was not very successful. I enrolled in a popular online course on AI a few years ago with Purdue Global and Simplilearn. What I failed to do with my newly acquired knowledge was build projects and practice. Just like anything new we learn, piano, riding a motorcycle or ballet dancing, practice makes perfect.

I decided to give AI another chance and enrolled in the MIT 12 Week program, MIT online Data Science and Machine Learning: Making Data Driven Decisions. This course promised to teach me the fundamentals of data science and machine learning, and how to apply them to real-world problems. I was curious to see if this course would be different from my previous experience, and but also tackling it with a different sense of purpose, resolve and determination.

What I Learned

The MIT 12 Week program was a challenging but rewarding experience for me. I learned a lot of new skills and concepts that helped me understand the basics of data science and machine learning. Some of the main topics that I learned were:

-- Data visualization and exploration: I learned how to use various libraries and tools, such as Matplotlib, Seaborn, and Plotly, to create and customize different types of charts and graphs, such as histograms, scatter plots, box plots, and heat maps. I also learned how to use descriptive statistics and correlation analysis to summarize and explore the characteristics and relationships of the data.

-- Data preprocessing and feature engineering: I learned how to clean and prepare the data for machine learning, such as handling missing values, outliers, and duplicates, encoding categorical variables, scaling numerical variables, and splitting the data into training and testing sets. I also learned how to create and select new features from the data, such as using polynomial features, interaction terms, and dimensionality reduction techniques, such as principal component analysis (PCA).

-- Machine learning algorithms and models: I learned how to use various supervised and unsupervised machine learning algorithms and models, such as linear and logistic regression, k-nearest neighbors, decision trees, random forests, support vector machines, k-means clustering, and hierarchical clustering. I also learned how to use various libraries and frameworks, such as Scikit-learn, TensorFlow, and Keras, to implement and train these models.

-- Model evaluation and improvement: I learned how to use various metrics and methods to evaluate and compare the performance and accuracy of the models, such as confusion matrix, accuracy, precision, recall, f1-score, roc curve, and auc score. I also learned how to use various techniques to improve the models, such as cross-validation, grid search, and regularization.

-- Data-driven decision making: I learned how to use the data and the models to make informed and evidence-based decisions, such as choosing the best model for a given problem, interpreting the results and insights from the models, and communicating the findings and recommendations to the stakeholders.

What I Liked

There were many aspects of the MIT 12 Week program that I liked and appreciated. Some of them were:

The course content and structure: The course was well-designed and organized, with clear and concise lectures, videos, readings, and assignments. The course covered a wide range of topics and concepts, from the basics to the advanced, and provided a comprehensive and practical introduction to data science and machine learning. The course also balanced the theory and the practice, and provided many examples and case studies from various domains and industries, such as healthcare, education, finance, and social media.

The instructors and mentors: The instructors and mentors were knowledgeable and experienced, and they explained the concepts and techniques in a clear and engaging way. They also provided helpful feedback and guidance, and answered the questions and doubts that I had along the way. They also encouraged me to think critically and creatively, and to apply what I learned to my own projects and interests.

The learning community and support: The course also provided a supportive and collaborative learning environment, where I could interact and learn from other learners and peers from different backgrounds and locations. I could also access various resources and forums, such as Slack, GitHub, and Stack Overflow, where I could share my work, ask for help, and exchange ideas and tips.

What I Disliked

There were also some aspects of the MIT 12 Week program that I found challenging. Challenges force us to extend ourselves and grow. Some of them were:

The course pace and workload: The course was very intensive and demanding, and it required a lot of time and effort to complete. The course covered a lot of material in a short period of time, and it sometimes felt overwhelming and stressful. I had to manage my time and priorities well and balance the course with my other commitments and responsibilities of being a husband, father, son etc.

The technical difficulties and issues: The course also involved using various tools and platforms, such as Jupyter Notebook, Google Colab, and AWS, which sometimes caused technical difficulties and issues, such as installation errors, compatibility problems, and connection failures. I had to troubleshoot and resolve these issues, which sometimes took a lot of time and patience.

The technical difficulties were a positive, because it forced me to dig deep and get comfortable with being uncomfortable. I learned a lot more by having to troubleshoot, search, read, experiment, build again, test, rinse and repeat. That is part of what we must do on a daily basis as engineers and technology practioners.

It reminded me of working on my master’s degree and figuring out that you must learn how to learn on your own. I was also able to draw on that knowledge in systems engineering and apply a lot of that methodology to the course and my approach

Conclusion

In conclusion, the MIT 12 Week program was a valuable and enjoyable learning experience for me. I learned a lot of new skills and concepts that helped me restart my AI journey a second time, and I gained more confidence and interest in data science and machine learning. I also faced some challenges and difficulties, but I overcame them with the help of the instructors, mentors, and peers. I would recommend this course to anyone who wants to learn or improve their data science and machine learning skills, and who is willing to invest the time money, and effort required.

My biggest takeaway is that I now need to go out and build ore hands on projects before the learnings become stale and fade away. Stay tuned as I begin to build out my project portfolio over the next several months. I plan to share my lessons learned.

Final thought, the internet can be our oyster where we can acquire a tremendous amount of knowledge. We just have to seek it out, vet it, then apply it to what we are trying to learn. For someone who grew up with libraries, encyclopedias and the dewey decimal system, I think I am navigating it all just fine. Smile !

Respectfully written,

Bennie

bbj