Part I
I started the data science (DS) Bootcamp with Practicum by Yandex during the first week of October 2021. I choose to attend a DS Bootcamp because I want to prepare myself well for a career in DS. The cohort and community manager, Ana, is always very responsive and supportive of any questions during the learning process. The instructors and tutors are very knowledgeable, and they actively answer any questions that I have within 24 hours and usually within only a few hours. The interaction on the Slack channels is very smooth, responsive, and engaging. If some tough questions are hard to explain by typing in Slack, tutors will happily hold 1-1 Zoom sessions with any questions. I found the personalized support very motivating and beneficial.
The DS Bootcamp has 17 modules/sprints on essential data science topics: basic python, data preprocessing, exploratory data analysis (EDA), statistical data analysis, integrated project 1, data collection and storage (SQL), introduction to machine learning, supervised learning, machine learning in business, integrated project 2, linear algebra, numerical methods, time series, machine learning for texts, computer vision, unsupervised learning, and the final project. Each module/sprint has chapters to study and interactive exercises and lessons. At the end of each module/sprint, an assigned sprint project requires applying the knowledge and skills in that module and some previous modules to solve problems. It is a cumulative learning process. Modules and projects in later modules require the skillful application of skills learned from previous modules. That also means that as one progresses, the project feels harder. So if you start off feeling, "Wow, these are too easy for me!" Then wait for a while, you will realize, "Wow, these are not that easy for me!" I like to be challenged and grow, so the progression of the complexity of the modules/sprints feels excellent!
It is a flexible learning process. There are deadlines for each module and project. A certain level or amount of extension is possible if one falls behind. If you want to finish faster, that is possible too. As long as you can complete the modules' lessons and projects and pass the project code reviews, you can continue and move forward. Speaking of project reviews, the reviewers are sharp at picking out mistakes. The perk is that they also offer more efficient code or approach to solve problems if my code is too bulky or output looks too "ugly." The community manager will coordinate the process and facilitate your needs, whether you are going fast or slow.
Now, 16 weeks into the program, I am at module 11. I spend a few hours learning and flexing my DS muscles each day. I am feeling stronger. This journey with many like-minded friends, peers, and data scientists is super enjoyable. I like the journey more and more. My heartfelt appreciation goes to everyone I come across on this journey. Thank you. 😊💖📚
Part II
Today, my integrated project two was finally approved in the early afternoon! 👏 This project took me seven days (about 4 hours each day) from start to finish, and it was tricky to handle on several aspects:
- The number of attributes from the datasets is a lot, and it is overwhelming to check them.
- The questions are hard to answer, and it took me a while to understand the project description, the technological process, and how to answer the questions.
- The code implementation was somewhat challenging, for example, using custom metrics in cross-validation and merging data with that many columns!
Now I can happily move on to the project for the linear algebra sprint. Yay! 🙌