BEGIN:VCALENDAR VERSION:2.0 PRODID:-//HasGeek//NONSGML Funnel//EN DESCRIPTION:Machine Learning\, Deep Learning and Artificial Intelligence: concepts\, applications and tools. NAME:Anthill Inside Miniconf – Pune REFRESH-INTERVAL;VALUE=DURATION:PT12H SUMMARY:Anthill Inside Miniconf – Pune TIMEZONE-ID:Asia/Kolkata X-PUBLISHED-TTL:PT12H X-WR-CALDESC:Machine Learning\, Deep Learning and Artificial Intelligence: concepts\, applications and tools. X-WR-CALNAME:Anthill Inside Miniconf – Pune X-WR-TIMEZONE:Asia/Kolkata BEGIN:VEVENT SUMMARY:Introduction to HasGeek\, Anthill Inside DTSTART;VALUE=DATE-TIME:20171124T043000Z DTEND;VALUE=DATE-TIME:20171124T044000Z DTSTAMP;VALUE=DATE-TIME:20210227T100624Z UID:session/ CREATED;VALUE=DATE-TIME:20171114T070029Z DESCRIPTION:\n LAST-MODIFIED;VALUE=DATE-TIME:20171114T070029Z LOCATION:Room 1 - Venue 1\, IN ORGANIZER;CN="Anthill Inside" BEGIN:VALARM ACTION:display DESCRIPTION:Introduction to HasGeek\, Anthill Inside in Room 1 in 5 minute s TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Analytics without paralysis! DTSTART;VALUE=DATE-TIME:20171124T044000Z DTEND;VALUE=DATE-TIME:20171124T050500Z DTSTAMP;VALUE=DATE-TIME:20210227T100624Z UID:session/ CATEGORIES:Crisp Talk,Beginner CREATED;VALUE=DATE-TIME:20171110T061538Z DESCRIPTION:Analytics gets adopted when decision makers are positively inf luenced by data. But is the corporate world looking at data like this? Are analytics teams telling their stories to grab Decision maker’s attentio n! Can they afford not to? Analytics doesn’t need you to solve a technic al problem but a “business” problem. And the only way to increase anal ytics adoption is to story tell. When presenting ideas to decision makers\ , realize that it is your responsibility to sell – not their responsibil ity to buy. Stories are the best way to influence!Data has to climb out of a dashboard & tell a story.\n\nAnd now AI & Machine learning are changing the way we can storytell. Today technology can work in tandem with human creativity to provide data-driven\, factual and interactive context to a s tory.\n\nIn this talk I will look at examples of how data insights can lea d to embedding analytics into the fabric of a company.\n\n### Speaker bio\ n\n\n LAST-MODIFIED;VALUE=DATE-TIME:20171114T070016Z LOCATION:Room 1 - Venue 1\, IN ORGANIZER;CN="Anthill Inside" URL: s-without-paralysis-Y19yPQBmYLeVfeRpWPVUiG BEGIN:VALARM ACTION:display DESCRIPTION:Analytics without paralysis! in Room 1 in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:(Not so) Straight (!) fun with Linear Regression DTSTART;VALUE=DATE-TIME:20171124T050500Z DTEND;VALUE=DATE-TIME:20171124T054500Z DTSTAMP;VALUE=DATE-TIME:20210227T100624Z UID:session/ CATEGORIES:Full talk,Beginner CREATED;VALUE=DATE-TIME:20171108T064533Z DESCRIPTION:We'll test the well known concept of Linear Regression using a live experiment!\nWe may chance upon 'feature engineering' and 'multiple linear regression' as we pass by.\n\n### Speaker bio\n\nI am a programmer with an odd love for maths. I enjoy simplifying heavy math protein into mo re absorbable amino acids\, only to be assimilated into plump biceps of co nfidence\, to be flexed when the situation demands.\nI want to infect peop le with the addictive epiphanies from solving math problems.\n\nand btw\, I have been working as a programmer on Data Science projects for the last 6+ years and as a programmer for last 13+ years.\n LAST-MODIFIED;VALUE=DATE-TIME:20171115T102816Z LOCATION:Room 1 - Venue 1\, IN ORGANIZER;CN="Anthill Inside" URL: traight-fun-with-linear-regression-SNs1bd1sxXN66gZq5T7w6X BEGIN:VALARM ACTION:display DESCRIPTION:(Not so) Straight (!) fun with Linear Regression in Room 1 in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Break DTSTART;VALUE=DATE-TIME:20171124T054500Z DTEND;VALUE=DATE-TIME:20171124T060500Z DTSTAMP;VALUE=DATE-TIME:20210227T100624Z UID:session/ CREATED;VALUE=DATE-TIME:20171030T091714Z DESCRIPTION:\n LAST-MODIFIED;VALUE=DATE-TIME:20171114T070012Z ORGANIZER;CN="Anthill Inside" BEGIN:VALARM ACTION:display DESCRIPTION:Break in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Bayesian methods in data analysis\, an introduction DTSTART;VALUE=DATE-TIME:20171124T060500Z DTEND;VALUE=DATE-TIME:20171124T064500Z DTSTAMP;VALUE=DATE-TIME:20210227T100624Z UID:session/ CATEGORIES:Full talk,Beginner CREATED;VALUE=DATE-TIME:20171108T064626Z DESCRIPTION:1. Start with basics of bayesian methods\, few historical anec dotes about the multiple interpretations of probability.\n 2. Cover pract ical examples and problem statements which are best analysed with bayesian methods.\n 3. Show some live coding examples using open source governmen t datasets from fields like econometrics or agriculture or healthcare.\n 4. Scratch the surface about algorithmic implementations: how the famous ' markov chain monte carlo' MCMC methods work.\n 5. Quick review of librari es/tools (pymc).\n 6. If you are excited with the idea\, how can you stud y further?\n\n### Speaker bio\n\nI work as head of data science at onlines\, an advertising technology startup based out of Pune. I have 7+ y ears of experience in data science and started in the field before it was a buzzword :-P. I have built multiple products\, handled consulting assign ments and delivered solutions using machine learning\, R and Python. I hol d a Master’s degree in Operations Research from Indian Institute of Tech nology\, Mumbai. \n\nBayesian methods have been my area of interest for a long time. Over the years\, I have formed few opinions about their usefuln ess and tried my best to understand the underlying theory\, that I would l ike to share through this talk.\n LAST-MODIFIED;VALUE=DATE-TIME:20171114T070010Z LOCATION:Room 1 - Venue 1\, IN ORGANIZER;CN="Anthill Inside" URL: -methods-in-data-analysis-an-introduction-TxyCm5HRScztmdng9RrjYV BEGIN:VALARM ACTION:display DESCRIPTION:Bayesian methods in data analysis\, an introduction in Room 1 in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Machine Learning in Molecular Biology DTSTART;VALUE=DATE-TIME:20171124T064500Z DTEND;VALUE=DATE-TIME:20171124T072000Z DTSTAMP;VALUE=DATE-TIME:20210227T100624Z UID:session/ CATEGORIES:Full talk,Beginner CREATED;VALUE=DATE-TIME:20171110T061517Z DESCRIPTION:1. Brush up on high-school biology.\n2. Introduction to some o f the new biotechnologies that produce data.\n3. Mixture models and why fe ature selection is important in an unsupervised learning kind of a setting \, with an example.\n4. An example of a Biological problem than can be for mulated as supervised learning.\n5. Some pictures of genetically modified creatures from our collaborators (that show machine learning works!).\n\n# ## Speaker bio\n\nI am part of a group of scientists at the National Chemi cal Laboratory\, Pune\, who use mathematics and computation to understand diverse aspects of Biology. I am a computer scientist by training and work primarily on designing probabilistic models as well as algorithms to lear n them\, all with the hope of solving fundamental problems in genomics.\n LAST-MODIFIED;VALUE=DATE-TIME:20171114T070008Z LOCATION:Room 1 - Venue 1\, IN ORGANIZER;CN="Anthill Inside" URL: learning-in-molecular-biology-XndG5A88cuwVizFJztab2v BEGIN:VALARM ACTION:display DESCRIPTION:Machine Learning in Molecular Biology in Room 1 in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Applications of ML in Ad Tech and Lifecyle of a ML project DTSTART;VALUE=DATE-TIME:20171124T072000Z DTEND;VALUE=DATE-TIME:20171124T080000Z DTSTAMP;VALUE=DATE-TIME:20210227T100624Z UID:session/ CATEGORIES:Full talk,Beginner CREATED;VALUE=DATE-TIME:20171114T065852Z DESCRIPTION:i. Intro\n 1. Bio\n 2. ML\n 3. PubMatic\n 4. IIMB\nii. Lif ecycle of Machine Learning project\n 1. Understanding Problem Statement\n 2. Research - Understanding Industry\, Domain and Field of study\n 3. C ollecting Data\n 4. Understanding and preparing data\n 5. Feature Select ion and Imputation\n 6. Data Sampling\n 7. Hypothesis testing and Descri ptive\n 8. Model Building\n 9. Tuning and Validation\n 10. Presenting t he Model\n 11. Deployment and Verification\niii. Conclusion and key takea ways\n\n### Speaker bio\n\nA Machine Learning/AI and Distributed Systems e ngineer who enjoys solving complex problems and design application and sys tems to work at scale.Have worked on engineering various complex projects which include building predictive ML project for online advertising\, deri ving interseting insights on IPL(Indian Premier League)\, building connect ors to offload data to Hadoop and even modifying Hadoop HDFS source code t o make Namenode more scalable. I have B.Tech in Computer Science from VIT\ , Pune and have specialization in "Big Data Analytics" from IIM Bangalore. \n LAST-MODIFIED;VALUE=DATE-TIME:20171115T090441Z LOCATION:Room 1 - Venue 1\, IN ORGANIZER;CN="Anthill Inside" URL: e-of-machine-learning-project-with-a-case-study-BmyZ7bWdaeVpGfg6gpVESo BEGIN:VALARM ACTION:display DESCRIPTION:Applications of ML in Ad Tech and Lifecyle of a ML project in Room 1 in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Lunch DTSTART;VALUE=DATE-TIME:20171124T080000Z DTEND;VALUE=DATE-TIME:20171124T090000Z DTSTAMP;VALUE=DATE-TIME:20210227T100624Z UID:session/ CREATED;VALUE=DATE-TIME:20171030T091820Z DESCRIPTION:\n LAST-MODIFIED;VALUE=DATE-TIME:20171115T082415Z ORGANIZER;CN="Anthill Inside" BEGIN:VALARM ACTION:display DESCRIPTION:Lunch in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:How similar are two pieces of text? A moderately broad and deep di ve in one of the fundamental topics in NLP. DTSTART;VALUE=DATE-TIME:20171124T090000Z DTEND;VALUE=DATE-TIME:20171124T094500Z DTSTAMP;VALUE=DATE-TIME:20210227T100624Z UID:session/ CATEGORIES:Full talk,Intermediate CREATED;VALUE=DATE-TIME:20171113T064930Z DESCRIPTION:1. Text Similarity\na. Definition and scope\n2. Application Ar eas\na. Information retrieval\nb. Paraphrase detection\nc. Natural languag e inference\nd. Plagiarism detection\n3. Types of Similarity\n4. Technique s\na. Supervised\ni. Classical techniques\nii. Deep neural network based t echniques\nb. Unsupervised \ni. Lexical\nii. Semantic\n5. Automatic Short Answer Grading\na. Context and motivation\nb. Word-similarity based techn iques\ni. Wisdom of students\nc. Siamese LSTM-based supervised ASAG techni que\n6. Conclusion\n\n### Speaker bio\n\nShourya Roy is the Head and Vice President of American Express Big Data Labs (BDL) which he took up in Dece mber 2016. In this role he is responsible for establishing and executing t he technical agenda for BDL working closely with the broader Decision Scie nce community and business units. Shourya is leading a team of scientists and engineers in the areas of machine learning\, artificial intelligence\, deep learning and cloud computing.\n\nPrior to joining American Express\, Shourya spent nearly fifteen years in the labs of IBM and Xerox playing s everal leadership roles in technical research\, research and strategic man agement\, customer facing business development. Shourya has a proven track record of conceptualize and initialize (by influencing business group lea ders)\, design and develop (by participating and leading research teams) a nd transfer (with software development partners) innovation from research labs to real life operations and business. \nShourya’s technical experti se spans Text and Data Mining\, Natural Language Processing\, Machine Lear ning\, and Big Data in which he is a well-known thought leader in several communities. His work has led to more than 60 publications in premier jour nals and conferences. He has been granted about 15 patents while tens of p atent applications are currently in different stages of patent lifecycle. He is an active member of the ACM and ACL communities - as a part of which he has been associated with multiple conference and workshop organisatio ns.\nShourya holds Ph.D.\, Masters and Bachelors Degrees in Computer Scien ce from IISc Bangalore\, IIT Bombay and Jadavpur University respectively. Shourya also has an MBA from Faculty of Management Studies (FMS)\, Delhi University.\nBeyond work Shourya is passionate about meeting and knowing p eople as well as following and playing multiple sports.\n LAST-MODIFIED;VALUE=DATE-TIME:20171115T082421Z LOCATION:Room 1 - Venue 1\, IN ORGANIZER;CN="Anthill Inside" URL: lar-are-two-pieces-of-text-a-moderately-broad-and-deep-dive-in-one-of-the- fundamental-topics-in-nlp-BiZQ67QZTrGW4pvJdbi57j BEGIN:VALARM ACTION:display DESCRIPTION:How similar are two pieces of text? A moderately broad and dee p dive in one of the fundamental topics in NLP. in Room 1 in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Applied Machine Learning for realtime #FairPlay against Fraud [spo nsored] DTSTART;VALUE=DATE-TIME:20171124T094500Z DTEND;VALUE=DATE-TIME:20171124T100500Z DTSTAMP;VALUE=DATE-TIME:20210227T100624Z UID:session/ CATEGORIES:Flash talks,Intermediate CREATED;VALUE=DATE-TIME:20171120T175626Z DESCRIPTION:1. Challenges at Dream11\, India's largest fantasy sports plat form\n2. Referral and promotional events\, user registration and game play .\n3. User data collection and preparing training data\n4. Regression and Gradient Boosted Models\n5. Scaling up for real-time decision making\n6. B usiness impact and key takeaways\n\n### Speaker bio\n\nAditya Prasad Naris etty is a Sr. Data Scientist @Dream11 building data driven products from f raud prevention\, User & Revenue estimation\, marketing attribution\, data pipelines and real-time M/L intelligence. Earlier\, he was heading the Da ta Science team at Craftsvilla building recommendation systems\, Data Plat form\, Search\, Autosuggestion\, real-time inventory profiling\, and Fashi on Recognition using CNNs.\n\nHe's an avid speaker in the Mumbai machine l earning community presenting at GDG Mumbai'17\, AWS conf'16\, DataNativesX \, HYSEA IIT-H\, Mumbai AI meetup and a couple of other meetups in Mumbai. \n\n\n LAST-MODIFIED;VALUE=DATE-TIME:20171124T070132Z LOCATION:Room 1 - Venue 1\, IN ORGANIZER;CN="Anthill Inside" URL: machine-learning-for-realtime-fairplay-against-fraud-Jz4Z5PCjLbf7L98o826cN e BEGIN:VALARM ACTION:display DESCRIPTION:Applied Machine Learning for realtime #FairPlay against Fraud [sponsored] in Room 1 in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Inference in Deep Neural Networks DTSTART;VALUE=DATE-TIME:20171124T100500Z DTEND;VALUE=DATE-TIME:20171124T104000Z DTSTAMP;VALUE=DATE-TIME:20210227T100624Z UID:session/ CATEGORIES:Full talk,Intermediate CREATED;VALUE=DATE-TIME:20171109T073020Z DESCRIPTION:- Intro DL Networks.\n - How do typical Deep Learning Archi tetures look. \n - A small section using example of one CNN and one L STM on what mathematical operations do they perform.\n- Advancements in Ha rdware\n - Intel Knight CPU's\n - Nervana \n - Volta GPU's\n- How exactly the operations are done on garden-variety hardware\n - SIM D\n - SIMT\n - GeMM\n- Different type of Architectures \n - CP U and GPU's\n - How do these work and bottlenecks\n- Role Played b y Memory access in speeds\n - How a lot of times memory is the bottlen eck instead of Compute\n- Changes in algortihms made to utilise these func tionalities\n - Example of Google's Inception V3 model\n - Two dif ferent type of RNN's\n- Advice\n - How to make your model more efficie nt at inference. \n - Some practical examples\n\n### Speaker bio\n\nSa urabh has been working at MAD Street Den\, Chennai as a Machine Learning E ngineer since past year and a half\,specifically working on Deep Learning based products. He loves to train Convolutional Neural Networks of all typ es and sizes for different applications. Apart from CNN’s he has special interest in recurrent architectures and discovering their powers. When he is not working on DL based stuff\, he loves to play around with micro-con trollers.\n LAST-MODIFIED;VALUE=DATE-TIME:20171120T175645Z LOCATION:Room 1 - Venue 1\, IN ORGANIZER;CN="Anthill Inside" URL: e-in-deep-neural-networks-HV82W5wjkoXkR63VUjWYYk BEGIN:VALARM ACTION:display DESCRIPTION:Inference in Deep Neural Networks in Room 1 in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Break DTSTART;VALUE=DATE-TIME:20171124T104000Z DTEND;VALUE=DATE-TIME:20171124T110000Z DTSTAMP;VALUE=DATE-TIME:20210227T100624Z UID:session/ CREATED;VALUE=DATE-TIME:20171030T091928Z DESCRIPTION:\n LAST-MODIFIED;VALUE=DATE-TIME:20171120T175649Z ORGANIZER;CN="Anthill Inside" BEGIN:VALARM ACTION:display DESCRIPTION:Break in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Doing Data Science on Cloud DTSTART;VALUE=DATE-TIME:20171124T110000Z DTEND;VALUE=DATE-TIME:20171124T114500Z DTSTAMP;VALUE=DATE-TIME:20210227T100624Z UID:session/ CATEGORIES:Full talk,Intermediate CREATED;VALUE=DATE-TIME:20171114T101034Z DESCRIPTION:Data scince on Cloud:\nImportance of running DS on Cloud?\nOpt ions for running ML on cloud platform\n - Using native co mpute and storage only.\n - Hosted Data platfrom\n - Machine Learning Services\n - Congnitive API Services.\nDemo : Using Cognitive Services of Google Cloud Platform- GCP vision API.\nOptions for running scalable DS models on Cloud:(Advantage\, Disadvantage\, Pricing)\n - AWS\n - Azure Machine learning\n - Google Cloud ML \nOther providers: IB M bluemix vs vs Domino datalab vs Datajoy\nDemo: Running DS model s using Tensorflow on Google Cloud ML(Using GPUs).\n\n### Speaker bio\n\n* Swapnil is right now contributing to Schlumberger Data Science team apply ing analytics in field of Oil and Natural Gas.Prior to this he was part of Snapdeal Realtime Analytics team as Lead Enginner.\nSwapnil in the past h as worked as Cloudera Trainer.He belives in learning and sharing his learn ing across the community.A frequent speaker in meetups and active presente r in conferences.\nWith more than 8+ years of experience\, Swapnil has con tributed in Domains of BFSI\,Ad Serving and eCommerce with Hadoop\,Spark a nd GCP as primary tech stack.\nPast conferences & Meetups:\n- https://expe\n- me-processing-and-watermarks-using-google-pub-su\n- http://www.bigdatainno\n- Dr Dobbs conference-B angalore- April 11-12\,2014\n\n* Ekansh Verma is right now working with Sc hlumberger Data Scince team as Data scientist.He has done his Bachelors\, Biomedical Engineering from IIT Chennai.He has good understanding of Deep Learning concepts. His primary expertise lies in Image classfication.\n LAST-MODIFIED;VALUE=DATE-TIME:20171120T175650Z LOCATION:Room 1 - Venue 1\, IN ORGANIZER;CN="Anthill Inside" URL: ta-science-on-cloud-J7jtcWKebZuWQTQ2yw42fm BEGIN:VALARM ACTION:display DESCRIPTION:Doing Data Science on Cloud in Room 1 in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Build intelligent\, real-time applications using Machine Learning DTSTART;VALUE=DATE-TIME:20171124T114500Z DTEND;VALUE=DATE-TIME:20171124T122000Z DTSTAMP;VALUE=DATE-TIME:20210227T100624Z UID:session/ CATEGORIES:Full talk,Intermediate CREATED;VALUE=DATE-TIME:20171114T065808Z DESCRIPTION:* Discuss the current-state-of-affairs for deploying Machine L earning models\n* Discuss shortcomings of this approach\n* Discuss the val ue of streaming data\n* Brief introduction to Apache Kafka and Streaming a pplications\n* Discuss how to use Apache Kafka to use ML models in real-ti me\n* Demonstrate how we use a Demography Prediction model in real-time\n\ n### Speaker bio\n\nJayesh leads the Personalisation team at Hotstar. He h as been building streaming applications using Apache Kafka for the last 4 years. At Hotstar\, the personalisation team builds Machine Learning model s for its 150 million users and delivers it real-time. He can be reached o n Twitter at @jayeshsidhwani\n LAST-MODIFIED;VALUE=DATE-TIME:20171120T175653Z LOCATION:Room 1 - Venue 1\, IN ORGANIZER;CN="Anthill Inside" URL: telligent-real-time-applications-using-machine-learning-M2aHPk8FzHqThz3v2p Fnzn BEGIN:VALARM ACTION:display DESCRIPTION:Build intelligent\, real-time applications using Machine Learn ing in Room 1 in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT END:VCALENDAR