Outsourcing Live Operations and bringing new life into old games with #IA

France
Joined October 2016
ML researchers work with fixed benchmark datasets, and spend all of their time searching over the knobs they do control: architecture & optimization. In applied ML, you're likely to spend most of your time on data collection and annotation -- where your investment will pay off.
19
320
63
1,833
Best of 2019: How GoldenEye 007 set a new standard for video game AI ubm.io/2xPPOkR
matchable.io retweeted
Mobile games generating more than $5m yearly store revenue: 2016: 1200 2018: 1900 2019: 2200 Data @appannie
Politicians have blamed violence on video games for decades. Said one expert: “The data on bananas causing suicide is about as conclusive.” nyti.ms/2KzK0lQ
138
1,306
119
3,789
26 years ago… 👀
8
42
6
112
matchable.io retweeted
Using #AI to Make Knowledge Workers More Effective: dy.si/9KGYo => The Real Power of AI = Accelerated Intelligence Amplified... Assisted... Augmented... ———— #FutureOfWork #Automation #Analytics #DataScience #MachineLearning #DigitalTransformation
Most companies using data
4
304
10
1,338
matchable.io retweeted
Apple has cemented a turning point for the future of the web. In building products, the question is moving from “how much data can we gather” to “how little data can we gather” to deliver the same value proposition.
49
630
82
3,349
matchable.io retweeted
How to Deploy Machine Learning Models: The Ultimate Guide buff.ly/2wh5bSP
126
5
330
matchable.io retweeted
10-page (PDF) #DataScience Cheat Sheet covers basic concepts in probability, #statistics, statistical learning, #MachineLearning, #DeepLearning, #BigData frameworks, and SQL: dy.si/GXLH2 ———— #abdsc #AI #DataScientists #DataLiteracy #BeDataBrilliant
108
4
229
Where's Ouya?
7
1
28
A couple days ago another team asked me to speak about Bayesian data analysis. I decided that instead of doing a nuts/bolts of how to fit/use Bayesian models, I would describe "Bayesian analysis: The Good Parts". <potentially controversial thread>
16
188
18
686
matchable.io retweeted
Python or R? Trick question: Both! Check out our PyData video playlist of R talks, including: "A Beginner's Guide to Random Forests - R vs Python" by Linda Uruchurtu: youtube.com/watch?v=fFPsjpP5… #rstats #scikitlearn #sklearn Don't forget to subscribe for updates!
4
1
15
matchable.io retweeted
Would argue that collecting the right data is the single most important aspect of building ML based products.
A stark contrast between Machine Learning vs other forms of mathematical modeling is that ML models often don't model extreme corner cases very well, because data in those areas is rare. Gathering data in important areas is as important a skill as building fancy neural networks.
7
41
10
190
When the client says "We don't need research, we know our users. 🧐" #UXProblems #UX #Research #Design #productdesign #designthinking 🤓 @FabienSerny @benjaminloembet @GregoirePoulain
8
277
38
423
Amazon built an AI to rate job applications. It analyzed 10 years of (male dominated) hires. Then it started penalizing resumes that included the word “women’s,” downgrading graduates from all women's colleges, and highly rating aggressive language. reuters.com/article/us-amazo…
108
2,237
884
2,998
Every so often I read a tweet about F2P that makes me sad for the developer sharing it. So I'm going to hit every common misconception I've seen in a single thread, and hope it helps somebody someday. Buckle up!
2
73
7
160