2015-8-1 Sklearn, XGBoost,等可重现数据

【基于Theano的可扩展深度学习框架deepy】"deepy: Highly extensible deep learning framework based on Theano" GitHub:O网页链接

【开源:基于Theano的CNN实现(dropouts/adagrad/momentum/max-margin layer/...)】O网页链接

【可重现数据驱动研究平台REP】全称是Reproducible Experiment Platform，统一封装TMVA, Sklearn, XGBoost, Uboost等分类实现，进行大数据集共享一致性对比试验，可在集群上完成并行训练 GitHub:O网页链接REP(Reproducible Experiment Platform)文档:O网页链接

【开源:Scikit-Learn兼容的(Python)半监督学习框架】"Semi-supervised learning frameworks for Python" GitHub:O网页链接

【高效的Python数据分析框架Ibis】O网页链接GitHub:O网页链接通过IPN了解Ibis:O网页链接Slide:《Ibis: Scaling the Python Data Experience》O网页链接云:O网页链接

【(Python)深度学习框架/库/工具汇总介绍】《Frameworks and Libraries for Deep Learning》Theano/Pylearn2/Blocks/Keras/LasagneO网页链接

【数据可视化框架/库/软件大列表】"Awesome dataviz"O网页链接

【(Python)三行代码实现Hinton's Dropout】《Hinton's Dropout in 3 Lines of Python - How to install Dropout into a neural network by only changing 3 lines of python》by TraskO网页链接

【GoogLeNet类可视化】《Visualizing GoogLeNet Classes》O网页链接GitHub(DeepDraw):O网页链接

【开源:(Python)方便的"One Pass"统计/回归计算库RunStats】"RunStats: Computing Statistics and Regression in One Pass"O网页链接

【幻灯:(PyData 2015)机器学习系统观】《PyData 2015 Keynote: "A Systems View of Machine Learning"》by Joshua BloomO网页链接云:O网页链接

【开源:(Python)马尔可夫链蒙特卡罗(MCMC)绘图包】"Python package to plot MCMC samples"O网页链接