Pycon tw 2013
- 1. Report for Pycon tw 06/08/2013 Wako.R showyou
- 2. Datamining Engineer Work ● Maintain Hadoop ● Maintain BI tool ● Hive, Pig, ... PRML reading circle Like Python Who am I?
- 3. ● Talk about foreign Conference ● Not talk about R, but talk about Data science ● Let Japanese makes slide in English The aim of report
- 5. Talk about: 1. Python 3(difference from Python2) 2. Algorithm 3. Collect data Keynote
- 6. cross validation sklearn.cross_validation Python for Data scientist Training CV f=kx
- 7. Boosting ensemble.GradientBoostingRegressor? example: Mean for a data set which contains outlier data set -> sampling some data points, calculate mean(=a learner) ... do N times Pycon for Data scientist
- 8. http://www.slideshare.net/jimmy_lai/big-data-analysis- in-python Bigdata with python
- 9. Python has many libraries for scientist. Scipy, Numpy, Scikit, Pandas, Disco - Lightweight Map Reduce NLTK - Natural Language ToolKit Pandas - Dataframe StatsModel - data structure Matplotlib - Plot method of R NetworkX http://networkx.github. io/documentation/latest/gallery.html Bigdata with python(2)
- 10. ● Stuff of Disqus ● Sentry https://getsentry.com/welcome/ ● Redis Scale
- 11. ● R is good language, but Python become statistical/machine-learning programing language ● Especially, a half of session talked about numpy, data manipulation, etc... ● a half of session shows the slide written in English ○ In Japan, how many people? My impression
- 12. Talked about pycon tw 2013 from datamining engineer view. "Numpy" Conclusion