Creative Talks: Data Pipeline - Fast Preprocessing For Complex AI
Time & Location
About The Event
The second Creative Talks of this fall will guide you through 𝘾𝙧𝙚𝙖𝙩𝙞𝙫𝙚 𝘿𝙤𝙘𝙠’𝙨 𝙙𝙖𝙩𝙖-𝙙𝙧𝙞𝙫𝙚𝙣 𝙙𝙚𝙨𝙞𝙜𝙣 𝙥𝙞𝙥𝙚𝙡𝙞𝙣𝙚. From statistical learning to deep neural networks, from manual preprocessing and labeling to estimating the price of your property.
>>>>
3 interconnected talks will cover the following topics:
𝙃𝙤𝙬 𝙩𝙤 𝙘𝙧𝙚𝙖𝙩𝙚 𝙧𝙚𝙡𝙞𝙖𝙗𝙡𝙚 𝙖𝙣𝙙 𝙥𝙧𝙚𝙘𝙞𝙨𝙚 𝙙𝙖𝙩𝙖 𝙛𝙤𝙧 𝙨𝙪𝙥𝙚𝙧𝙫𝙞𝙨𝙚𝙙 𝙡𝙚𝙖𝙧𝙣𝙞𝙣𝙜
> Tasks that require both manual and automated production of labels for supervised learning algorithms (links to online data sources, grocery store data, etc.)
> Methods and processes that help us label data in a very efficient way while achieving high accuracy, at the lowest possible cost
𝙊𝙗𝙩𝙖𝙞𝙣𝙞𝙣𝙜 𝙙𝙖𝙩𝙖 𝙛𝙧𝙤𝙢 𝙥𝙪𝙗𝙡𝙞𝙘𝙡𝙮 𝙖𝙫𝙖𝙞𝙡𝙖𝙗𝙡𝙚 𝙨𝙤𝙪𝙧𝙘𝙚𝙨
> How we can help banks decide if a company is eligible for loan
> Our algorithm to find company data in publicly available sources, consisting of two parts: matching and scraping
> Searching for links to company FB pages, websites, …
> Scraping relevant data and using these to rate and score the company
𝙋𝙧𝙞𝙘𝙚 𝙢𝙖𝙥𝙨: 𝙝𝙤𝙬 𝙢𝙪𝙘𝙝 𝙞𝙨 𝙮𝙤𝙪𝙧 𝙝𝙤𝙪𝙨𝙚 𝙬𝙤𝙧𝙩𝙝?
> Automating the mortgage process
> Estimating the prices of houses
> Using various public data sources – from historical real estate data to data from the Czech cadastre and Czech statistical office
>>>>
𝙎𝙥𝙚𝙖𝙠𝙚𝙧𝙨:
> Matúš Ondreička, data engineer
> Adam Hanka, data analyst
> Jan Pavel & Matěj Pacovský, data analysts
>>>>
Small refreshments will be provided.
Entrance is free but please don’t forget to register.
Looking forward to seeing you at Creative Dock HQ!