You are on page 1of 18

Name- Pratyush Ranjan

SAP ID- 500091264


Batch 1 AIML (non-hons)

Machine learning lab


Experiment-1
Weka

What is weka?
Weka is a collec on of data mining machine-learning algorithms. Weka
includes data pre-processing, classi ca on, regression, clustering,
associa on rules, and visualisa on tools. It is bene cial to see examples of
data preprocessing, model building, and result interpreta on. Weka is not
at all di cult to use; it is simple and straigh orward.
It is open-source so ware that anyone can use for free; it assists other
people from various branches such as electrical, mechanical, and civil who
get IT jobs and must work on di erent algorithms, as they do not have
enough deep knowledge to use this tool at rst. Though it is not
recommended for a developer to use this, it can help them throughout
the process.
An overview of how weka works are-
ti
ffi

ti
ft

ti
ff
fi

ti

tf
fi
fi

ti
As shown in the preceding diagram, the rst and most important step in
using Weka is to collect primary or secondary data and upload it to the
tool. The data must then be preprocessed, and Weka provides numerous
tools for doing so. Then you'll see various ML algorithm op ons and can
choose one based on the results you want. The di erent models can be
used on the same dataset. You can then compare the outputs of various
models and choose the best one for your needs.
Steps how to use weka-
Step 1- Prepare the dataset, for a reference we are downloading any csv
le from the internet.
Step 2- Open the weka tool you will see this window.
fi

fi

ff

ti

Step 3- Click on explorer.

The following window will open-


Step 4- click on open le, there will be a whole list of le extensions will
come, you can choose any type of le that you want to.
fi
fi
fi

A whole list of les present in your device will appear.

You would choose one of the alterna ves, such as Classify, Cluster, or
Associate, based on the type of machine learning model you are
a emp ng to construct. With the help of the a ributes selec on,
tt
ti
fi
ti
tt

ti

features can be automa cally chosen to provide a smaller dataset.

The sta s cal output of the model processing would be provided by


WEKA. You can examine the data using the visualisa on tool it o ers.
ti
ti
ti
ti
ff

Experiment-3
Introduc on to python libraries

ti

Experiment-5
Logis c regression
ti

Experiment 4
Implementa on of Mul ple Linear Regression

ti

ti

Experiment 12
Skewness and Kurtosis

You might also like