Python Libraries for Data modelling.
Well I have been using python for the past two years or so in my ML projects and I'll be sharing with you some amazing libraries and tools that are going to get you up and running those projects in machine learning.I wouldn't want to go into so much details but this will give you the information that you need to get started.
Here are the most important packages that when it comes to Machine learning.
1.Pandas
It is a library built in the python language that we use in data analysis and manipulation
The name pandas comes from two words 'panel data' and 'econometrics' well for those of us who've not yet been to a statistics class econometrics is as a term that is used to refer to a set of data items from indivduals in the same population. It is also the name of a gigantic animal native to Southern China.
Pandas has a number of amazing features and functionalities including:
~Data restructuring and reshaping tools.
~Time series and analysis techniques.
~Linear regression tools.
~Row and column insertion options for large data sets well even for smaller ones.
~Data sets merging and joining and other manipulation tools.
Besides all this pandas has the flexibility of allowing you to work with quite a number of data formats i.e CSV ,SQL ,Excel and even JSON.
You may have come across a very common phrase know as multi dimensional arrays either in a math class or in a blog. Or even came across some disturbing matrices while doing some calculations.
Well the good news is that a solution to all intensive statistical computing is Numpy.
Originally created by Jim_Hugunin Numpy now an open sources software with contributors from across the globe provides
Several functionalities of dealing with multidimensional arrays and objects.
Among this with a rich source of data sets you can create sub arrays from existing data, slice and perform indexing on the same arrays.
Since arrays are stored in the same memory space unlike their counter parts lists and also its vast efficiency in working with latest tech this makes it incredibly fast in accessing and working with data sets.
In addition to being crossplatgorm, numpy can work with hundreds of mathematical functions including;
~Trigonometric functions
~Sums products and differences .
~Exponents.
~Hyperbolic functions and many more that you can Read here.
I usually refer of to as a "cousin" of Numpy heavily relied on Numpy while being developed but has some amazing tools when it comes to -Numerical integration
-Optimization techniques.
-Linear Algebra.
-And statistics.
Also one of the most used libraries in data science, has a good documetion which can be accessed on their website and an amazing community of developers.
Besides being the best in Image processing tools it is also good in Classification, Regression, clustering, data preprocessing and many more.
Developed and managed by Google, it is by so far the best machine learning library I have ever come across .With numerous tools and awesome support for beginners, you can do anything really with this amazing library.
It can be used in building Natural Language processing tools ,Speech recognition, building neural networks and deploying any Ml model that you can really think of .
It is the industry's best when it comes to Machine Learning.
You would agree with me that this not the end of the list but, for visualization and presentation you can also use the following libraries;
2.Pydot
3.Seaborn(based on Matplotlib)
4.Bokeh
Let talk on this and many more in my subsequent posts.
Comments
Post a Comment