Your First Step-by-Step Machine Learning Project

I’ve been learning machine learning for a few months now, and the hardest part isn’t the math — it’s knowing where to start. Most tutorials assume you already know what you’re doing. This post is for people like me who just want a simple, working example they can run and modify.

Here’s a step-by-step walkthrough using the classic Iris dataset. All you need is Anaconda and a Jupyter notebook.

Setup

1. Download and install Anaconda
2. Launch Jupyter Notebook
3. Copy the code below into cells and run them

Load the libraries

import pandas
from pandas.tools.plotting import scatter_matrix
import matplotlib.pyplot as plt
from sklearn import model_selection
from sklearn.metrics import classification_report
from sklearn.metrics import confusion_matrix
from sklearn.metrics import accuracy_score
from sklearn.linear_model import LogisticRegression
from sklearn.tree import DecisionTreeClassifier
from sklearn.neighbors import KNeighborsClassifier
from sklearn.discriminant_analysis import LinearDiscriminantAnalysis
from sklearn.naive_bayes import GaussianNB
from sklearn.svm import SVC

This gives you everything: pandas for data handling, matplotlib for plotting, and scikit-learn for the actual ML models.

The beauty of this example is that you’re comparing six different classifiers on the same dataset — logistic regression, decision trees, KNN, LDA, naive Bayes, and SVM — and you can see which one performs best. No theory required. Just run it, see the results, and then start asking why.

That’s how I’m learning: code first, theory second. If something works, I dig into why. If it doesn’t, I dig into why not. The Iris dataset isn’t going to change the world, but it’s a solid foundation to build on.