Analytics India Magazine’s hackathon platform MachineHack recently concluded its hackathon titled “Visualisation is Beautiful: Uber City Traffic Visualization Challenge”. We talked to the winners of the hackathon to know about them and to know how they solved this month’s mind-bender.
Surenther Has Rank 1 On The Leaderboard
Surenther is into data science since 2016. He was in the data science and analytics centre of Karpagam College of Engineering, Coimbatore. He says that data visualisation is the heart of data science as it gives solutions as the visualisation tools like Tableau is making easy for him. Tableau made him understand the keynotes of visualisation and he shares his skills in visualization that he acquired, with his students.
Approach to solving the problem:
After having a look at the data, Surenther realised that he needs more dimensions to make charts, so he carefully handled the data carefully making sure that the values do not change. He then moved to visualisation of the charts, data repentance time series to make an easy forecast and said that the single dashboards gave him more information.
Experience on MachineHack:
Talking about his experience on the platform, he says that it is a good platform to get started in data science hackathons. He thanked MachineHack for the opportunity to participate in the hackathon.
Sahil Mendhiratta Has Rank 2 On The Leaderboard:
Mendhiratta says that a data science career involves a mix of skills around data handling, modelling and presentation skills. He started his career towards data science in 2018 when he was introduced to it in his very first organisational project. He developed a deep interest in it and started learning it at a good pace. He read various blogs on data science and data analytics and realised that it is a hot technology that will create a boom in the market in the coming years. Courses on EDX, DataCamp, Coursera helped him develop his foundation. He believes that data science will definitely bring a drastic change in every industry to solve real-world problems.
Approach to solving the problem:
According to Mehdhiratta, the data provided was the kind a BI developer actually wants, in a numerical format. Although he had to do some transformations and data cleansing operations on the data. The major dimension to show on dashboards was date and time and so he focused primarily on showing time span at various hierarchies of date and time, like a year, day and month level. He also focused on analysing traffic at different point of times. He used the platform of Microsoft PowerBI because it would give highly customised and beautiful charts giving deeper insights view.
Experience on MachineHack:
It was a great learning experience with MachineHack and it helped him grow, enrich knowledge and develop confidence in data science. He said that he would look forward to participating in the upcoming hackathons of MachineHack.
Sreyan Ghosh Has Rank 3 On The Leaderboard
Ghosh is a third-year undergraduate student from Christ University, Bangalore from the department of computer science and engineering. His data science journey began in his second year of engineering when he attended an introductory machine learning workshop organised by one of his seniors in college. Ghosh said that he went back home the same day and signed up for Andrew Ng’s machine learning course on Coursera and he continues with various such courses even after.
Ghosh said that being a student, he did take some time and determination to learn the statistics and mathematics side of data science but he never gave up. He also received a lot of encouragement from his teachers in his department. He also co-founded a club in his college called Neuron which has been growing exponentially since its inception.
The first ever competition that he took part in was on Kaggle where the job was to predict poverty levels in Costa Rica. Since then I kept taking part in several online and offline competitions, which helped him to stay in touch with all new innovations in the data science world and I learnt a lot more quickly than just taking up courses online. One thing that he says has learnt so far in his data science journey is that data science is not just about making complicated models, it’s about taking better decisions.
Approach to solving the problem:
Talking about how he solved the hackathon problem, Sreyan said that he was the last one to submit the solution. He initially saw the leaderboard filled with visualisations from Power BI and he knew that he needed something different to stand out. So he used Plotly tool of Python. The best thing about Plotly, he says, is the customisation that it allows. Custom functions can be written that can be used to present data in exactly the way you want it to be. Moreover, the vibrant colours that can be used to catch attention immediately.
Initially, after cleaning the dates column which had dates in different formats, he had started with the visualization part. The most important thing was to choose wisely between all the columns in the dataset, the columns which could be used to portray the most out of the data. He primarily used the ‘mean’ features rather than the ‘upperbound’ and ‘lowerbound’ features for making his visualisations. Some of his key visualisations included a violin plot of the mean travel time across days of a month, days of the week, boxplot of mean travel time across years and across months of a year and distribution plots of mean, upper-bound and lower-bound travel times across different parts of a day. The whole process of selecting the best visualisations to put up on MachineHack Ghosh said was quite iterative. He also took the help of my friends and asked them to review it before he could select the 6 best visualisations. In the process, he discarded about 50 visualisations before he came up with my 6 best visualisations on the eve of the submission deadline.
Experience on MachineHack:
Ghosh said that I was a great learning experience at MachineHack and competing with top contenders, especially professional data scientists can make students and aspiring data scientists grow and
See where they stand in terms of industry standards. Machine Hack provides an amazing competitive environment for professionals and data science enthusiasts. Ghosh also said that he will keep taking part in hackathons on MachineHack and is looking forward to more hackathons. He also thanks the HOD of his department, Dr Balachandran K and his teacher who supported him in his data science journey. He also thanked his friends who helped spread the word and all those who voted for him.
MachineHack recently concluded one more hackathon called “Predict A Doctor’s Consultation Fee Hackathon” and we published the winners’ article on how they solved the hackathon problem at Analytics India Magazine. MachineHack hosts many interesting hackathons and also has interesting prices for the winners. It has recently launched a new hackathon called “Predict The Flight Ticket Price Hackathon” to solve the unpredictability of flight ticket prices.
