DBTB INT Chester Chen r

[Music] sure Chester chin I work for alpine data and the director of engineering well the part exciting is that you can see any different applications and a different techniques and a different tracks and so in the past and we went attending at the other different conferences you don't see so many different particular verticals you know you have certain type of us machine learning specific talks about it today this time they bring many verticals and in particular politics and all the others genomics and you know it's very interesting [Music] today is cool he's a new bait it was called a new beggin ya so you know data bring a lot of her reach information so i think for enterprise and they have been setting up a big data infrastructure for over a long time but many of them haven't really explored the potentials of the data and how they use them and how they change the business and with the recent years a special last of the three four years and people kept to the point that i say the tools the infrastructures have to mature to a point that they can explore this in the lever genus so many of many of the businesses that have make decisions that you know if we they couldn't leverage in this like the other big companies does and then they may you know they don't have the competitive advantage over other competitors and so so this is a become really really getting faster and faster especially with the recent development in the AI world [Music] the meaning society is so my talking's are really talking about how to do with a real time the visualizations are for machine learning I think even though there's a you know a lot of advances into a machine learning algorithms that are sometimes in the machine learning take a long time to run and we have no clue what's inside that you know I so the cost functions that being flat hold in the weather the machine learning convergence are not and there's no monitoring tools or for that so so the talk I'm gaming is really illustrate that for a certain part of the algorithms it's actually quite easy to modify this and you can actually directly plug into your you know application with the same techniques and also on I'm also hoping you know by showing the techniques I've done with alpine and hoping that these spar communities can directly put these interface and hooks into the spark interfaces so to enable other applications monitoring visualizations and into a spark I don't think I have one because I'm not data science myself so but but just observing to the other power our critical workers and have several co-workers who have done a lot of work and thus far communities and it becomes parking meters I felt that of course you have you have that you need to learn mathematics so basically a needler machine learning theories and all that stuff and also you have to be willing to hands on you know learning spark and it's a great one and by you know contributing to a spark in other open source communities and actually help you to to learn the big data infrastructures as well as the machine learning outcomes and by doing more and they essentially back and you know getting your into our familiars not just the theory but also the practice of the machine learning and of course on top of that it besides implementing them you have to use you know basically finally use cases and apply them in attending meet up zone is another way to learn [Music] you