Machine learning basic library abstract
Before starting machine learning I’m starting machine learning required library using python .
- pandas
- Numpy
- matplotlib
- sklearn
The Pandas Library:-
To install pandas use
“pip install pandas”
pandas is used for handle dataframe and is used to load data using different functions like
pd.load_csv()
pd.load_html()
pd.load_sql()
or different functions these all things in pandas loaded for dataset.
when we analyse dataset we use drop data,fillna data or imputer these function used for get perfect dataset for training .
2. The Numpy Library:-
At the dawn of the Python language, the developers began to need to perform numerical calculations, especially when this language began to be considered by the scientific community.
The first attempt was Numeric, developed by Jim Hugunin in 1995, which was successively followed by an alternative package called Numarray.
Both packages were specialized for the calculation of arrays, and each of them had strengths depending on which case they were used.
Thus, they were used differently depending on where they showed to be more efficient. This ambiguity led then to the idea of unifying the two packages and therefore Travis Oliphant started to develop the NumPy library.
Its first release (v 1.0) occurred in 2006. From that moment on, NumPy has proved to be the extension library of Python for scientific computing, and it is currently the most widely used package for the calculation of multidimensional arrays and large arrays.
In addition, the package also comes with a range of functions that allow you to perform operations on arrays in a highly efficient way and to perform high-level mathematical calculations.
Currently, NumPy is open source and licensed under BSD. There are many contributors that with their support have expanded the potential of this library.
Numpy is used for array.
when we required to convert image into array we use numpy function.
in deep learning CNN model all image converted in vector and these vector stored into arrays
to install numpy :
On Linux (Ubuntu and Debian)
sudo apt-get install python-numpy
On Linux (Fedora)
sudo yum install numpy scipy
On Windows with Anaconda
conda install numpy
3.The Matplotlib Library
Matplotlib is a Python library specializing in the development of two-dimensional charts (including 3D charts);
in recent years, it has been widespread in scientific and engineering circles (http://matplolib.org).
Among all the features that have made it the most used tool in the graphical representation of data, there are a few that stand out:
- extreme simplicity in its use
- gradual development and interactive data visualization • expressions and texts in LaTeX
- greater control over graphic elements
- export in many formats such as PNG, PDF, SVG, and EPS.
Matplotlib is designed to reproduce as much as possible an environment similar to Matlab in terms of both graphic view and syntactic form
matplotlib is used for
visualization.
for plot of result in regression algorithm uses matplot.lib for understanding output
4.The scikit-learn Library:-
scikit-learn is a Python module that integrates many of machine learning algorithms.
This library was developed initially by Cournapeu in 2007, but the first real release was in 2010.
This library is part of the SciPy (Scientific Python) group, a set of libraries created for scientific computing and especially for data analysis, many of which are discussed in this book.
Generally these libraries are defined as SciKits, hence the first part of the name of this library.
The second part of the library’s name is derived from Machine Learning, the discipline pertaining to this library.
