Classic open source data sets for machine learning

In the table below are some of the most commonly used datasets compiled by Layman that can be used throughout machine learning. These datasets also frequently appear in official examples for SkLearn, Spark ML, and Tenserfolw.

Extraction was done by Barry Becker from the 1994 Census database. A set of reasonably clean records was extracted using the following conditions: ((AAGE>16) && (AGI>100) && (AFNLWGT>1)&& (HRSWK>0))

Prediction task is to determine whether a person makes over 50K a year.

This data is extracted from the 1994 US census database and can be used to predict whether a household’s income exceeds 50K$/year. The class variable of the dataset is whether the annual income exceeds 50K $, and the attribute variable contains important information such as age, job type, education background, occupation, race, etc. It is worth mentioning that there are 7 category variables among the 14 attribute variables.

The 20 Newsgroups data set is a collection of approximately 20,000 newsgroup documents, partitioned (nearly) evenly across 20 different newsgroups.

The data set contains about 20,000 newsgroup documents, evenly distributed among 20 different newsgroups, and is a classic data set for text classification, which is a popular data set for experiments in text applications of machine learning techniques, such as text classification and text clustering.

MNIST data set Is a data set for handwriting recognition in the field of machine learning. The data set contains 6 10,000 training sets and 10,000 sample test sets. , the width and height of each sample image is 28*28. The size of these data sets has been normalized and formed a fixed size, so the pre-processing work has basically been completed. In machine learning, many mainstream machine learning tools (including Sklearn) use this data set as an introductory level introduction and application.

For obtaining the iris dataset, we use the API provided by Sklearn instead of downloading the dataset by ourselves.

1. Data acquisition and description

import pandas as pd

from sklearn.datasets import load_iris



data = load_iris()

df = pd.DataFrame(data.data, columns=data.feature_names)

df.info()



# info describes the result



<class 'pandas.core.frame.DataFrame'>

RangeIndex: 150 entries, 0 to 149

Data columns (total 4 columns):

sepal length (cm)    150 non-null float64

sepal width (cm)     150 non-null float64

petal length (cm)    150 non-null float64

petal width (cm)     150 non-null float64

dtypes: float64(4)

The memory usage is: 4.8 KB

Copy the code

2. Data examples

df.head()

Copy the code

3. Data description

Describe data. You can use describe to describe various dimensions of a data set, such as the total number of dimensions, average values, etc.

df.describe()

Copy the code

This is just a quick example, but if you want to go deeper, you can check out the API details on the official website.

The UCI dataset consists of more than 400 data sets for supervised and unsupervised learning, many of which are referenced repeatedly in numerous other data tools, such as Iris, Wine, Adult, CarEvaluation, Forest Fires, and more.

Address: archive.ics.uci.edu/ml/

• www.zhihu.com/question/63…
• zhuanlan.zhihu.com/p/25138563
• Blog.csdn.net/gzhermit/ar…