Correlation One

I was a total noob when I started the program and through the program, I have been able to get a very thorough overview of data science processes. From learning to use pandas module for exploratory data analysis to using amazon relational databases and tableau, I have experienced how data science projects evolve from a mere idea to a well-formulated research question to final conclusions. I could always reach out to the TAs, mentors, and professor Pillai himself for any confusion. The correlation one team was also always there for help. In fact, they were sending me personalized internship recommendations until I told them that I had committed to another internship. 
Like many things though, it is what you make of it. The possibilities are endless if you capitalize on the opportunities available throughout and after the program. I highly recommend this program for any data enthusiast regardless of your previous experience. I felt that the program has something for everyone. 

This is a great option if you want to learn data science that is applicable to a lot of different fields.

Disclaimer: Some months after I graduated with honours in 2019, I applied to be a teaching assistant in Correlation One, and in early 2021, I started working as a full time product designer with them. My main job is making sure we iterate in the right direction to make the course more accesible, fun and engaging. Our current programme contains a lot more stuff than when I was a student.
From DS4A as a student, my very first impression was positive because the case studies went from marketing, to genomics, to public transportation, to many other subjects. The curriculum covers basically the whole range of techniques used in modern data science. Some of the subjects you'll find here (notice this is not an exhaustive list):

• EDA (Exploratory Data Analysis.) Stuff I liked: Missingness plots, grouped boxplots, strip plots.
• Data cleaning.
• NLP (Natural Language Processing.) This was my favourite subject and actually one of the most useful professionally so far.
• Dashboarding. Tools: Dash (you'll like it if you're a programming buff) and Google Data Studio, which has a nice graphical interface.
• Causality. I'm a causality nerd, so I found it really nice learning how to actually run a statistical matching algorithm. I loved this and Simpson's paradox.
• Reinforcement learning, Neural Networks, classification, and Computer Vision. I'm not really into AI, but if you like that, you will definitely learn a good deal of useful stuff.
• Hypothesis testing. 
• Linear and logistic regression. I'd never heard of L1 or L2 regularisation, or the Bonferroni correction. This was pretty cool. 
• K-means.
• Geographical data. This was my second favourite subject. It was really exciting plotting interactive maps!

I'm a political scientist and social researcher, not a computer scientist or an engineer, so I'm less concerned about the latest tools or algorithms and more about the way I can interpret data correctly. Even despite that, I'd say the programme was very accesible and the way subjects were taught was pretty intuitive and just made sense. I had some previous coding experience, but there were certainly lots of things about programming that I learnt in this bootcamp as well. The language of choice was Python, in Jupyter, with heavy reliance on the Pandas library, but if you come from a different language, it will be easy to adapt —many of my classmates came from R and Java and they did fine. I can definitely say that I came out being a lot more conscious about how to not misinterpret data, which was the main thing I wanted to get from this programme.

DS4A Colombia 1.0 was an amazing  learning experience in Data Science.
We  Learn by doing and teamwork. 

• Team of 5 people, balanced and multidisciplinary. 
• The team can build an  data folio, report, and codebase in a few weeks. (Final project and Datathon ).
• The week assignments helps you a lot for embrace the knowledge learned.
• Build a community, new friends and  whatsapp / slack team buildings.
• This program was unique and best are  in class ; a lot exchange knowledge, a  mix of feelings,  happiness, satisfaction and competition.  Helps you to growth at professional and personal level.
• Share LinkedIn and job postings
• DS4A certification in Colombia provides value to get a job.
• Practice English listening, writing, speaking, and presenting.
• Amazing C1  teachers and TAs. Natesh Pillai is a great teacher from Harvard  We can ask and interact with them. 
• Content  to with real cases / datasets and walkthrough about technology, databases, algorithms and programming techniques  structured to ramp up and reach a clear final goal.  
• Real problem cases with business context. Feedback and Review from TAs.

Data Science for All is the perfect intro to data science boot camp for anyone who has interest in the field but has little knowledge of what it is or what data scientists do! The cases, instructors, and mentors all work together to make the field more accessible by providing you with a foundation for entry. Not to mention the cohort being full or people from all walks of life, industries, and professions at various stages of their careers. Correlation One did an amazing job at creating a community that will stand the tests of time. The courses are completely free and still of outstanding value. You exit the program with an understanding of what you know and don't know that you can use to guide further learning and capitalize on skills that will get you to that data job. You also leave the program with a strong network of like minded individuals and a data science project you can be proud of and add to your portfolio. I can not recommend it enough. If you are thinking of applying, you should!

13 Saturdays, with some Saturdays off
10am -3pm EST for lectures and case studies
3pm-8pm EST for group work

Data Science for All was an amazing experience that I was fortunate enough to participate in. Considering that it was the first cohort to enter the program, the fact that there were only a few problems with the program is a testament to the dedication and forethought that the program leaders have for this endeavor.

NOTE:
Since we were the first cohort, we were the guinea pigs, and we tried different things during the program. Certain aspects might have and will change (for example, they now have a python self-paced mini boot camp before the actual program, as opposed to during the first few weeks of the program), so please refer to the program website and leaders to find the most up-to-date information.

MY BACKGROUND:
I graduated with a math degree in 2017, and I took two intro-to-computer-science and a data structures course. However, I forgot most of what I was taught, and spend most of 2020 trying to relearn everything, with an emphasis in python. It was during this time that I learned about DS4A.

APPLICATION PROCESS
I don't remember the application process that well but what I can say the following:

• The initial application didn't take too long. They're interested in why you want to join the program, and what prior experience/education you have. They also want to know more about your experience as an underrepresented individual.
• The assessment involves a lot of statistical terms that I wasn't used to. However, in context you can understand what the question is asking you and what the different responses mean. A review of statistical terms isn't necessary, but it wouldn't hurt. 
• The interview was a lot shorter than I thought it would be. They go over some of the questions that you answered in the initial application and they let you elaborate more. They check to see if you pass some minor technical requirements you might have (more on this later). They also ask about you're current employment status, and whether you'd be available to consider certain job opportunities that come through the program (note: You don't need to be unemployed or looking for a new job to enter the program. I knew multiple people who had no intention of leaving their job, but rather wanted to upgrade their skillset)

I heard roughly 2-3 weeks later that I had been accepted into the program.

TECHNICAL REQUIREMENTS
The only real technical requirement that we had for the program was that our computer have at least 8gb of ram. This is so they can use a program called Docker, which to put simply, sets up a mini computer within your current computer. You'll be running the software for the program through Docker.
Even if you didn't have 8gb of ram, people were still able to participate, but it just makes troubleshooting a bit hard in the long run.

THE INSTRUCTORS/ADMINISTRATORS (will talk about TA's and Mentors later)
The main instructor for the program is Professor Natesh Pillai, professor of Statistics at Harvard University. However, you'll know him better as DJ Tesh. Natesh is an absolutely wonderful and amazing individuals, and within seconds of hearing him speak, I was able to relax and I knew I was in good hands. Not only is he passionate about data science, but he's also passionate about the mission of DS4A and teaching. He brought a lot of energy and kept everyone engaged and focused. 
Sometimes other individual's would come in to teach certain parts of the program, and I felt really bad for them since they were always compared to the powerhouse that is Natesh. However, they were also great teachers, and tried their best to make sure that they explained everything as clear as possible. 
The rest of of Correlation One crew was also amazing. One thing that surprised me was not only the speed of their response, but also the sincerity they had. They worked with participants to make the program better and were very willing to admit and change mistakes they might have made.

SATURDAY CLASSES AND CASE STUDIES
Classes started at 7am PST, and usually had case studies from 7:10-ish to about 12 pst. Almost everything you'll learn will be through the case studies, which has real world situations. A majority of the case studies do a great job of teaching the concept at hand, and I would refer to them multiple times during the program. During the case study, if you had questions you could ask the presenter directly through the Slack channel or ask your TA's who were on standby to answer questions. Some of the concepts or tools we learned were

• Python
• Pandas
• SQL
• Data Cleaning
• Data visualization
• Hypothesis testing
• and much more!

One thing about the classes, though, was that the pacing was never perfect, and that because the skill level of all the participants varied widely. Some individuals had prior experience with coding, like myself, and so we could follow along well enough. Other with no coding experience struggled a bit during the first couple of cases. However, they are making changes to fix this, such as having a python bootcamp before the actual program. and experimenting with recording the lessons and having TA's go over the recording within their groups. Another Issue was the timing of some concepts. For example, I personally would have liked learning more about the statistical aspects of data science much sooner, since it would have helped with our project. However, I think this is a minor issues that can be easily fixed for future cohorts.
The rest of the day was usually filled with a mixture of another case study, some community event, an invited speaker sharing their experiences, TA meetings, or free time to work on your projects. One thing I'll say was that the speakers they had were amazing, and I think it's safe to say that everyone looked forward to hearing these individuals.

TAS AND MENTORS
Roughly 50 participants will be assigned to a pair of TA's and they're one of the first individuals you'll contact when you have questions or problems regarding the material or your project. They answer questions during the lectures and have office hours on Saturday and during the week to help you out with your studies. From what I heard from friends in different TA groups was that most if not all of the TA's were very helpful and friendly. I can say that my TA's were great, and one TA in particular went above and beyond when our group was having major problems with our project.
The Mentors are individuals who you'll contact roughly once every 2 weeks. They'll mentor you with regards to your project and your future career goals. From my personal experience our mentorship didn't go smoothly, but that more of an error on our part and due to circumstances outside of our control. I heard most mentorships were really great and that people were able to learn a lot from them.

ASSIGNMENTS AND THE PROJECT 
We had 3 assignments we had to turn in for my cohort. You'll get credit and receive you certificate if you at least make an effort to finish it, but in order to receive a certificate with distinction, you'll need to a score of 90(?) or more on at least 2 of the assignments.
These assignments were not easy, and I would highly advise against starting them the night before. However, they're not impossible to do. You might need to do a little extra research in order to answer certain questions or to use certain tools, but it wasn't anything impossible. You're also allowed to work with others (so long as you didn't copy their work word for word), so these assignments usually lead to creating friendships with people who were struggling along with you. 
The project was the most important part of the program, and in my opinion the hardest. You'll be working on it starting the first week or two, and roughly every other week you'll be turning in a certain part of the project. You'll be working with 4 other individuals, and the idea is that the 5 of you together will have all the skills and knowledge necessary to finish this project. The projects aren't graded, per say, but rather they want to see that you applied what you've learned and made an honest attempt to answer a question. 

JOB ASSISTANCE
During and after the program ends, there will be career fairs where a good amount of companies will present a bit about who they are, what they do, and how they're using data science in their organization. They'll share what positions they have available and would be open to questions. One issue multiple people had was that a lot of the organizations at first were looking for either people with a lot of experience or people with masters/PhDs. However, this is more of a problem with the companies that presented and less of a problem with Correlation One. They told us they were actively working to make companies lower or drop their educational requirements where possible and to give DS4A participants a chance, and there has been a notable difference from those first presentations to now. I'm hoping that as knowledge of who DS4A is and what their mission is, that even more opportunities will come.
They also have a platform called C1-connect, where you can share your portfolio and your job preferences, and during certain times of the year they would match companies with you to go through the application process. However, I haven't gone through the whole process yet, so I can't say much about it for now.

SUMMARY
This program is an amazing opportunity to learn about data science and enter the field, despite the few problems it has an especially considering that its free. It isn't easy and requires dedication, but by doing so you'll be better prepared for life after the program. It doesn't teach you everything about the data science field, but I feel that it gave me the head start as well as the tools to continue further in this path. 
I'm glad I applied, and I'm sure you will too.