QA Official

Naive Bayesian Classification (very naive explanation)

https://qaofficial.com/post/2019/04/17/70440-naive-bayesian-classification-very-naive-explanation.html 2019-04-17
Recently, I have been working on a short text classification system. I have tried to use Naive Bayesian algorithm to make a baseline model (this algorithm was not adopted in the end). Naive Bayesian method is a very simple and efficient classification algorithm. This algorithm has been watched several times intermittently. Today, I will make a summary. The contents refer to Li Hang's Statistical Learning Method and Goodfellow's Deep Learning.

Recent Information Class

https://qaofficial.com/post/2019/04/17/70418-recent-information-class.html 2019-04-17
Toutiao: 1. Intel Wireless Charging Dramatically Draws Down and Businesses Lay Off Staff;2.IDC: Global Mobile Phone Assembly Factory Shipment Ranking in the First Quarter;3. U.S. research found that cell phone use is associated with cancer.4. Zhejiang police destroyed the industrial chain of stolen and sold Apple mobile phones worth more than 100 million yuan.5. Finnish government accuses Microsoft of Nokia Von Trapped;6. CTS Technology in China Fines US$ 34.9 Million

Talk about my understanding of Microservices

https://qaofficial.com/post/2019/04/17/70392-talk-about-my-understanding-of-microservices.html 2019-04-17
Microservices is a very hot concept in recent years, especially in the context of the Internet, Microservices's theory has the opportunity to be widely practiced.However, in the process of practice, people's understanding of Microservices is indeed very different. How can we really master Microservices's architecture theory?Through this article, I would like to share with you my understanding of Microservice. What is Microservices The emergence of Microservices has brought many benefits

Version 1.0 of Term frequency–inverse document frequency Optimization Algorithm for News Tag Extraction-Based on jieba Word Segmentation

https://qaofficial.com/post/2019/04/17/70399-version-1.0-of-term-frequencyinverse-document-frequency-optimization-algorithm-for-news-tag-extraction-based-on-jieba-word-segmentation.html 2019-04-17
Version 1.0 of Term frequency–inverse document frequency Optimization Algorithm for News Tag Extraction-Based on jieba Word Segmentation declaration Reproduction is acceptable, but please indicate the source.Mark the original link and respect the author's hard work. Original Link This paper is written with reference to the paper "Based on Improved Keyword Extraction from Chinese Websites

Zhi-Hua Zhou Machine Learning Review Summary Chapter 3

https://qaofficial.com/post/2019/04/17/70375-zhi-hua-zhou-machine-learning-review-summary-chapter-3.html 2019-04-17
linear model basic form 线性模型试图学得一个通过属性的线性组合来进行预测的函数,即f(x)=w1x1+w2x2+…+wdxd+b,w和b学得之后,模型就得以

[furnace smelting AI] machine learning 002- mark coding method

https://qaofficial.com/post/2019/04/17/70444-furnace-smelting-ai-machine-learning-002-mark-coding-method.html 2019-04-17
[furnace smelting AI] machine learning 002- mark coding method [Python library and version number used in this article]: Python 3.5, Numpy 1.14, Scikit-Learn 0.19 There are various forms of marks in supervised learning. For example, marks for face recognition may be ["Little Red", "Little Flower", "Cui Hua"...], these marks are like gobbledygook for machine learning, so in order for machine learning to "understand" these marks, the marks of these text

brief description of penguin CMS function

https://qaofficial.com/post/2019/04/17/70410-brief-description-of-penguin-cms-function.html 2019-04-17
Full View of Penguin CMS Functions 文章模块 功能 说明 文章管理 支持文章的「发布、删除、移动、复制、排序、推荐、置顶、隐藏、显示」操作等 文章分类 支持多级分类 定时发布 支持

how to do deep learning based on Spark: from Mllib to keras, elephas

https://qaofficial.com/post/2019/04/17/70428-how-to-do-deep-learning-based-on-spark-from-mllib-to-keras-elephas.html 2019-04-17
Spark ML model pipelines on Distributed Deep Neural Nets This notebook describes how to build machine learning pipelines with Spark ML for distributed versions of Keras deep learning models. As data set we use the Otto Product Classification challenge from Kaggle. The reason we chose this data is that it is small and very structured. This way, we can focus more on technical components rather than prepcrocessing intricacies.

preprocessing data preprocessing in sklearn

https://qaofficial.com/post/2019/04/17/70480-preprocessing-data-preprocessing-in-sklearn.html 2019-04-17
1. sklearn preprocessing Standardization is standardization, which converts data into data with zero mean and one variance as much as possible, such as standard normal distribution (Gaussian distribution).In practice, we will ignore the distribution of the data, just change the mean value to centralize the data, and then divide the discontinuous features by their standard deviation. 1.1 Standardization: De-averaging, Variance Scaling Standardization standardization standardization: adjust the distribution of characteristic data

sklearn Data Preprocessing Normalization for Python Learning Notes

https://qaofficial.com/post/2019/04/17/70484-sklearn-data-preprocessing-normalization-for-python-learning-notes.html 2019-04-17
please refer to sklearn_ data processing API for help! Standard Normalization Normalized to a mean of 0 and a variance of 1 sklearn.preprocessing.scale function: standarddize ataset alonganyaxisFirst post the main source code, at first glance, very messy, in fact, look closely, is more than a few judgment sparse matrix such as conditional code. #coding=utf-8 import numpy as np from scipy import sparse def _handle_zeros_in_scale(scale, copy=True): ''' Makes sure that whenever