10 Key Takeaways From the Groupinar “Data Driven Innovation”
Data is a new source of inspiration. The Internet and mobile connectivity has enabled a plethora of data that offers unprecedented opportunities in many domains. Using automated analytical methods, it reveals patterns humans alone might never see and offers new approaches to age-old decision-making and problem-solving processes. From Data Science for X-ray Astronomy, which helps us unravel physical laws under the most extreme conditions known, to financial markets emanating massive amounts of data from which machines can, in principle, learn to invest with minimal initial guidance from humans, data has unleashed endless innovation.
At no other time in history have humans felt so impressed and intimidated by what machines can do. With individuals, organizations and governments gradually realizing how useful data can be, more interfaces and visualization tools are bubbling up to make sense of them.
Speakers of this Connectedreams’ groupinar, who are from diverse sectors like Academics, Bioinformatics, Non-profit, Finance and Internet, discussed how the data has immersed into every field and how analyzing of data has become crucial for any project, or as a matter of fact, even an idea!
In-case you couldn’t grab the live feed of the groupinar on here are 10 key takeaways for you. If you are still interested to watch the panelists speaking in more depth about the topic, here is the video:
https://www.youtube.com/watch?v=3vI1s-mIgic
Follow us on YouTube and Facebook to keep in track with our next live online Groupinar.
The Key Takeaways :-
1. Ashish spoke about the usefulness and challenges of data saying “With lot of data, it is challenging but it’s also very useful. The benefit of large data is that you don’t necessarily have to try the most sophisticated things. But in terms of challenges, the biggest problem is incremental learning. With this, the challenge is how do you maintain accuracy, and how do you react fast enough. One important thing is also to choose the right matrix.”
2. Prukalpa, while working with social cops talked about the challenges she faced while collecting offline data saying “What you’re doing is that you are effectively converting a human into a sensor, so the challenges involved there are that this person has never ever collected data before, this person has never used a mobile phone before, and this person doesn’t know English. So we had to actually change the language of our app”.
3. Sarosh spoke about the challenges faced by him managing the NLP system at twitter, he said “It’s quite hard one reason is because its only 140 characters, but come to think of it, the hardest to work with the fact that the language people use on twitter, is very noisy, it’s not regular English. So the traditional NLP techniques, that have been developed for under text formats like for articles don’t apply to twitter”.
4. Sarosh also said “Deep learning combined with big data has made this future leaning thing kind of absolute, so you no longer need to sit down and do any future engineering. But, the downside is that deep learning is pretty much a black box, using previous methods, you have a really good understanding of what exactly is happening, what the system is doing, how the algorithm operates, but, deep learning its really hard to look under the hood”.
5. Ashish spoke about where actually the value of deep learning is and said “For the data we can’t easily comprehend and break it down, like if we have a spreadsheet, we easily know what are the columns and what each column means, but when you are looking at a picture, you really have no idea how the brain is decoding the picture right. So that is where I think deep learning is helpful. Because let’s say we have a lot of training samples, you know what your objective is, you can just throw the deep learning at the vast amount of data and come out with an answer .So I think that is where the value of deep learning lies.”
6. Prukalpa, when asked about the type of people to hire for a startup like hers, she said “So one of the things we care about a lot is how badly does this person wants to solve a problem. The only thing that remains is the people and how badly you want to solve complex problem, how badly are you close to our mission, that’s to being able to empower the world’s most important decisions through data. And that is kind of the core of building any team and everything else builds its way up.”
7. Ashish also gave some personal advice, saying “There is a lot of hype about data driven decisions, but I wouldn’t say that for a new business, you should totally depend on that. If you are trying to figure out your career, trying to start a new company, you cannot replace your human intelligence with data. I just don’t want people to get caught up in the hype. Being able to understand where data is valuable and where data is not. That is a big part of anyone who gets into data science.”
8. Sarosh was asked on how the journalists are actually using the analysis, to which he answered “They actually write articles about it. So if you go on our websites, you can see articles being posted on Washington, and if you are at the outlets, they literally take the analysis, pass it to the journalist and then they take that down and turn that down into articles.”
9. A question which came from the audience was how Prukalpa is planning to connect to connect IOT with data, to which she said, “We actually are doing a lot with IOT. We have already been looking at connecting to incredibly low cost biometric devices or fingerprint scanners to be able to get more private information or to be able to get more authentic information. We have already connected to air pollution sensor.”
10. Thinking about the future of data, Ashish said “For people who want to pursue data science, obviously there is a lot of demand and it’s not going to go away. But in the future, it won’t become a specialization, it would become a requirement. In terms of AI in general, I am very excited! The thing that I am most excited about is being able to have a natural conversation with the computer, with AI.”
Connectedreams.com is an online data-driven, cultural context-aware networking platform, bridging the role model & mentorship gap. Request an Invitation here.
