Representation learning——
principal component analysis
1. Summary
Principal component analysis was proposed by Karl Pearson in 1903 and improved from another perspective by Harold Hotelling in the 1930s. It is a method suitable for high-dimensional data and exploratory data analysis. As one of the first algorithms applied to pattern recognition and machine lea ...
Added by twilitegxa on Sat, 25 Dec 2021 07:17:37 +0200
[machine learning sklearn] Decision Tree algorithm
Tip: here is a hard Ken 📕 If Xiaobai makes a mistake, change it. Don't spray it:)
preface
The sky can be mended, the sea can be filled, the mountains can be moved, the sun and the moon are past, and can not be pursued again.
Decision Tree is based on tree structure to make decisions. (classification, regression) A decision tree contai ...
Added by deniscyriac on Sat, 25 Dec 2021 00:36:26 +0200
NumPy multidimensional array, several functions to create the array
To create an array object:
You can create an array of ndarrays through the array function of the NumPy library. Generally speaking, ndarray is a common data container, that is, all elements in it need the same type. NumPy library can convert data (list, tuple, array or other sequence types) into ndarray array
1. Create an array object using a ...
Added by Cetanu on Tue, 21 Dec 2021 20:26:56 +0200
matplotlib: Art Brush sees the world
matplotlib: Art Brush sees the world
1, Overview
1. Three layer api of Matplotlib
The principle or basic logic of matplotlib is to render graphics on the canvas with Artist objects. It is similar to the steps of human painting:
Prepare a canvas or drawing paperPrepare paint, brush and other drawing toolspaint a picture
[external chain pict ...
Added by CraigRoberts on Mon, 20 Dec 2021 21:48:40 +0200
Fundamentals of machine learning algorithms DAY 2
feature selection
Reason for feature selection
Redundancy: some features have high correlation and are easy to consume computing performance Noise: some features have a negative impact on the prediction results
What is feature selection
Feature selection is simply to select some features from all the extracted features as the features o ...
Added by Hatch on Thu, 04 Nov 2021 02:08:16 +0200
Machine learning sklearn ---- first knowledge of KMeans
summary
KMeans is an unsupervised learning method. He is a classification algorithm. It is used to explore the original data and classify the samples with the same attributes in the original data. This article only talks about the simple use of KMeans. For the content about the evaluation results, see my next article.
Several concepts in ...
Added by leafface on Sat, 16 Oct 2021 08:32:10 +0300
Use sklearn to compare various machine learning methods: SVM, XGBoost
Summary of common machine learning algorithms in sklearn (with python code) - Jianshu (jianshu.com)https://www.jianshu.com/p/641707e4e72c
Note: when using XGBoost algorithm, you need to set X_ train,y_ The data type of Tensor type variables such as train is set to nd array type, while other algorithms can directly use Tensor data type, ...
Added by keyoung on Mon, 11 Oct 2021 06:44:57 +0300
❤️ 20000 words, summarizing 50 pandas high-frequency operations [illustrated and worthy of collection] ❤️
Point, knock on the blackboard
First of all, this paper follows the traditional teaching, point to point! Only some functions or processing methods that are frequently used by individuals are introduced.The examples in this article are only used for demonstration. Generally, the examples do not modify the original data. If the code will modif ...
Added by Daukan on Tue, 21 Sep 2021 14:21:22 +0300