ai4r 1.13 → 2.0

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA256:
+  metadata.gz: 8d01465e193f7bda084a9cd76bc10e1dd4422b7c384eb0fe924d7a5f705123db
+  data.tar.gz: 64ffbe98250389997fbdc9fa1a93c4b79154c1b9d3cb4a5a6577c086ee392214
+SHA512:
+  metadata.gz: 2bf0626ab8227c93fbc8a784717806ec3c3b72a4e6d2f375f232be8ad0be9e5f1ee70c6dfc2be02cd2c6dd40311cb5ac1d840f3acc92300cc208c8ccfc839425
+  data.tar.gz: 5140c67b5213fa3ed1a9017c1d4dd09c05cba0786d27bc8b97c32e7d72d8638347f80ae22536ecacd4efb10742a88c2e7e263878baf73edcb064749564a1d8a0

data/README.md

@@ -0,0 +1,174 @@
+# AI4R — Artificial Intelligence for Ruby
+
+🎓 **Welcome to AI4R**
+
+Current version: 2.0
+
+Your Lightweight Lab for AI and Machine Learning in Ruby.
+
+AI4R isn’t just another machine learning library. It’s a learning playground. No black boxes and no bulky dependencies—just clean, readable Ruby implementations of core AI algorithms so you can explore, modify and really understand how they work.
+
+## Installation
+
+AI4R is distributed as a gem and requires Ruby 3.2 or later.
+
+Install the gem using RubyGems:
+
+```bash
+gem install ai4r
+```
+
+Add the library to your code:
+
+```ruby
+require 'ai4r'
+```
+
+## Where to start?
+- [Beginner Track](docs/learning_path_1_beginner.md) – Build core intuition for AI—step by step, in Ruby.
+- [Intermediate Track](docs/learning_path_2_intermediate.md) – From "I can run a model" to "I can tune, extend, and build smart stuff that actually works."
+- [Advanced Track](docs/learning_path_3_advanced.md) – Time to stop following recipes and start writing your own.
+
+
+## 🧭 What’s Inside?
+
+A quick map to AI4R’s built‑in toolkits, grouped by type. Each folder comes with examples and benchmark runners so you can dive right in.
+
+### 🤖 Transformers – *Play with the Building Blocks of Modern LLMs*
+
+> *“Meet your future coworker / overlord.”*
+
+This is not a full GPT—but it is the core logic, stripped down and readable.  
+AI4R ships with a bite-sized, dependency-free Transformer implementation that supports:
+
+- **Encoder-only** mode (like BERT)
+- **Decoder-only** mode (like GPT)
+- **Seq2Seq** mode (like T5)
+
+📂 Code: `lib/ai4r/neural_network/transformer.rb`
+Docs: [Transformer guide](docs/transformer.md)
+
+💡 **Try this**:  
+Load up the transformer and walk through a simple forward pass.  
+Everything from attention weights to layer normalization is short enough to read and understand in one go.
+
+### 🧠 Classifiers – Make Predictions
+"What’s the most likely outcome?"
+
+You’ll find in [lib/ai4r/classifiers/](lib/ai4r/classifiers/):
+
+- `ZeroR`, `OneR` – the simplest baselines
+- `LogisticRegression`, `SimpleLinearRegression`
+- `SupportVectorMachine`
+- `RandomForest`, `GradientBoosting`
+- `MultilayerPerceptron`
+
+Docs: [logistic_regression.md](docs/logistic_regression.md), [random_forest.md](docs/random_forest.md)
+
+Try this: run `compare_all.rb` to benchmark classifiers on real datasets.
+
+### 🔍 Clusterers – Find Hidden Patterns
+"What belongs together?"
+
+Includes in [lib/ai4r/clusterers/](lib/ai4r/clusterers/):
+
+- `KMeans`
+- `DBSCAN`
+- `Hierarchical` clustering variants
+
+Docs: [kmeans.md](docs/kmeans.md), [dbscan.md](docs/dbscan.md)
+
+Try this: cluster the Iris dataset using both KMeans and DBSCAN.
+
+### 🧬 Neural Networks – Learn From Data
+"What if we build a brain?"
+
+- Backpropagation – classic feedforward network
+- Hopfield – associative memory model
+- Transformer – a tiny GPT‑style block (encoder, decoder, seq2seq)
+
+Code: `lib/ai4r/neural_network/`
+
+Try this: open `transformer.rb` and trace each step—it’s short enough to grok in one sitting.
+
+### 🔎 Search Algorithms – Explore Possibility Spaces
+"What’s the best path?"
+
+You’ll find in [lib/ai4r/search/](lib/ai4r/search/):
+
+- `BreadthFirst`, `DepthFirst`, `IterativeDeepening`
+- `A*`
+- `MonteCarloTreeSearch`
+
+Docs: [search_algorithms.md](docs/search_algorithms.md)
+
+Try this: run A* and DFS on a maze and time the difference.
+
+### 🧪 Genetic Algorithms – Evolve a Solution
+"Let’s mutate our way to a better answer."
+
+- Generic GA framework
+- A Traveling Salesman Problem (TSP) chromosome
+
+Code: `lib/ai4r/genetic_algorithm/`
+
+Try this: tweak the mutation rate in the TSP example.
+
+### 🧭 Reinforcement Learning – Learn by Doing
+"Reward me, and I’ll improve."
+
+- Q‑Learning
+- Policy Iteration
+
+Code: `lib/ai4r/reinforcement/`
+
+Docs: `docs/reinforcement_learning.md`
+
+Try this: run a grid‑world training loop and watch the agent build its own policy.
+
+### 🕵️ Hidden Markov Models – Guess What’s Hidden
+"You can’t see the states—but you can infer them."
+
+Code: `lib/ai4r/hmm/hidden_markov_model.rb`
+
+Docs: `docs/hmm.md`
+
+Try this: model a weather prediction problem with hidden states and visible activities.
+
+### 🧠 Self‑Organizing Maps – Compress Dimensions
+"Can we project complex data onto a simpler map?"
+
+- Kohonen‑style SOM
+
+Code: `lib/ai4r/som/`
+
+Try this: reduce high‑dimensional vectors into a 2D neuron grid and color it based on class.
+
+## 🧪 Benchmarks: Experiment & Compare
+
+Each algorithm family has a benchmark runner:
+
+- `bench/classifier/`
+- `bench/clusterer/`
+- `bench/search/`
+
+Shared tools in `bench/common/` make it easy to run head‑to‑head comparisons, track runtime, accuracy and more, and output clean reports.
+
+Docs: `docs/benches_overview.md`
+
+Try this: run `bench/search/astar_vs_dfs.rb` and explain why A* usually wins.
+
+
+```bash
+git clone https://github.com/SergioFierens/ai4r
+cd ai4r
+bundle install
+ruby bench/classifier/compare_all.rb
+```
+
+
+
+## 💬 Feedback?
+
+This library is maintained for the joy of it (and perhaps a misplaced sense of duty to Ruby). You can do whatever you want with it—it’s unlicensed. If you build something cool or just find it useful, drop a note in the [project's comments](https://github.com/SergioFierens/ai4r/discussions).
+

data/examples/classifiers/hyperpipes_data.csv

@@ -0,0 +1,14 @@
+city,age,gender,marketing_target
+New York,25,M,Y
+New York,23,M,Y
+New York,18,M,Y
+Chicago,43,M,Y
+New York,34,F,N
+Chicago,33,F,Y
+New York,31,F,N
+Chicago,55,M,N
+New York,58,F,N
+New York,59,M,N
+Chicago,71,M,N
+New York,60,F,N
+Chicago,85,F,Y

data/examples/classifiers/hyperpipes_example.rb

@@ -0,0 +1,22 @@
+# frozen_string_literal: true
+
+require_relative '../../lib/ai4r/classifiers/hyperpipes'
+require_relative '../../lib/ai4r/data/data_set'
+
+# Use fully qualified class names instead of including modules.
+
+# Load the training data
+file = "#{File.dirname(__FILE__)}/hyperpipes_data.csv"
+data = Ai4r::Data::DataSet.new.parse_csv_with_labels(file)
+
+# Build the classifier using custom parameters
+classifier = Ai4r::Classifiers::Hyperpipes.new.set_parameters(tie_break: :random).build(data)
+
+# Inspect the generated pipes
+pipes_summary = classifier.pipes
+puts 'Pipes summary:'
+pp pipes_summary
+
+# Classify new instances
+puts "Prediction for ['Chicago', 85, 'F']: #{classifier.eval(['Chicago', 85, 'F'])}"
+puts "Prediction for ['New York', 25, 'M']: #{classifier.eval(['New York', 25, 'M'])}"

data/examples/classifiers/ib1_example.rb

@@ -0,0 +1,12 @@
+# frozen_string_literal: true
+
+require_relative '../../lib/ai4r/classifiers/ib1'
+require_relative '../../lib/ai4r/data/data_set'
+
+file = "#{File.dirname(__FILE__)}/hyperpipes_data.csv"
+data = Ai4r::Data::DataSet.new.parse_csv_with_labels(file)
+
+classifier = Ai4r::Classifiers::IB1.new.build(data)
+
+sample = ['Chicago', 55, 'M']
+puts "Prediction for #{sample.inspect}: #{classifier.eval(sample)}"

data/examples/classifiers/id3_example.rb

@@ -1,13 +1,15 @@
+# frozen_string_literal: true
+
 # Author::    Sergio Fierens
 # License::   MPL 1.1
 # Project::   ai4r
 # Url::       http://www.ai4r.org/
 #
-# You can redistribute it and/or modify it under the terms of 
-# the Mozilla Public License version 1.1  as published by the 
+# You can redistribute it and/or modify it under the terms of
+# the Mozilla Public License version 1.1  as published by the
 # Mozilla Foundation at http://www.mozilla.org/MPL/MPL-1.1.txt
 
-require File.dirname(__FILE__) + '/../../lib/ai4r/classifiers/id3'
+require_relative '../../lib/ai4r/classifiers/id3'
 
 # Load data from data_set.csv
 data_filename = "#{File.dirname(__FILE__)}/id3_data.csv"
@@ -17,13 +19,16 @@ data_set = Ai4r::Data::DataSet.new.load_csv_with_labels data_filename
 id3 = Ai4r::Classifiers::ID3.new.build(data_set)
 
 # Show rules
-puts "Discovered rules are:"
+puts 'Discovered rules are:'
 puts id3.get_rules
-puts 
+puts
 
 # Try to predict some values
-puts "Prediction samples:"
-puts "['Moron Sur (GBA)','4','[86 m2 - 100 m2]'] => " + id3.eval(['Moron Sur (GBA)','4','[86 m2 - 100 m2]'])
-puts "['Moron Sur (GBA)','3','[101 m2 - 125 m2]'] => " + id3.eval(['Moron Sur (GBA)','3','[101 m2 - 125 m2]'])
-puts "['Recoleta (CABA)','3','[86 m2 - 100 m2]'] => " + id3.eval(['Recoleta (CABA)','3','[86 m2 - 100 m2]',])
-puts "['Tigre (GBA)','3','[71 m2 - 85 m2]'] => " + id3.eval(['Tigre (GBA)','3','[71 m2 - 85 m2]',])
+puts 'Prediction samples:'
+puts "['Moron Sur (GBA)','4','[86 m2 - 100 m2]'] => #{id3.eval(['Moron Sur (GBA)', '4',
+                                                                '[86 m2 - 100 m2]'])}"
+puts "['Moron Sur (GBA)','3','[101 m2 - 125 m2]'] => #{id3.eval(['Moron Sur (GBA)', '3',
+                                                                 '[101 m2 - 125 m2]'])}"
+puts "['Recoleta (CABA)','3','[86 m2 - 100 m2]'] => #{id3.eval(['Recoleta (CABA)', '3',
+                                                                '[86 m2 - 100 m2]'])}"
+puts "['Tigre (GBA)','3','[71 m2 - 85 m2]'] => #{id3.eval(['Tigre (GBA)', '3', '[71 m2 - 85 m2]'])}"

data/examples/classifiers/id3_graphviz_example.rb

@@ -0,0 +1,17 @@
+# frozen_string_literal: true
+
+require_relative '../../lib/ai4r/classifiers/id3'
+
+# Load the training data
+file = "#{File.dirname(__FILE__)}/id3_data.csv"
+data_set = Ai4r::Data::DataSet.new.load_csv_with_labels(file)
+
+# Build the tree
+id3 = Ai4r::Classifiers::ID3.new.build(data_set)
+
+# Export DOT representation
+File.open('id3_tree.dot', 'w') { |f| f.puts id3.to_graphviz }
+puts 'Decision tree saved to id3_tree.dot'
+
+# You can also inspect the tree as nested hashes
+p id3.to_h

data/examples/classifiers/logistic_regression_example.rb

@@ -0,0 +1,11 @@
+require 'ai4r/classifiers/logistic_regression'
+require 'ai4r/data/data_set'
+
+items = [[0, 0, 0], [0, 1, 1], [1, 0, 1], [1, 1, 1]]
+labels = %w[x1 x2 class]
+set = Ai4r::Data::DataSet.new(data_items: items, data_labels: labels)
+
+reg = Ai4r::Classifiers::LogisticRegression.new
+reg.set_parameters(learning_rate: 0.5, iterations: 2000).build(set)
+
+puts reg.eval([1, 0])

data/examples/classifiers/naive_bayes_attributes_example.rb

@@ -0,0 +1,13 @@
+# frozen_string_literal: true
+
+require_relative '../../lib/ai4r/classifiers/naive_bayes'
+require_relative '../../lib/ai4r/data/data_set'
+
+file = "#{File.dirname(__FILE__)}/naive_bayes_data.csv"
+set = Ai4r::Data::DataSet.new.load_csv_with_labels(file)
+
+bayes = Ai4r::Classifiers::NaiveBayes.new.set_parameters(m: 3).build(set)
+
+puts bayes.class_prob.inspect
+puts bayes.pcc.inspect
+puts bayes.pcp.inspect

data/examples/classifiers/naive_bayes_example.rb

@@ -1,16 +1,15 @@
-require File.dirname(__FILE__) + '/../../lib/ai4r/classifiers/naive_bayes'
-require File.dirname(__FILE__) + '/../../lib/ai4r/data/data_set'
-require File.dirname(__FILE__) + '/../../lib/ai4r/classifiers/id3'
-require 'benchmark'
+# frozen_string_literal: true
 
-include Ai4r::Classifiers
-include Ai4r::Data
+require_relative '../../lib/ai4r/classifiers/naive_bayes'
+require_relative '../../lib/ai4r/data/data_set'
+require_relative '../../lib/ai4r/classifiers/id3'
+require 'benchmark'
 
-data_set = DataSet.new
-data_set.load_csv_with_labels File.dirname(__FILE__) + "/naive_bayes_data.csv"
+data_set = Ai4r::Data::DataSet.new
+data_set.load_csv_with_labels "#{File.dirname(__FILE__)}/naive_bayes_data.csv"
 
-b = NaiveBayes.new.
-      set_parameters({:m=>3}).
-      build data_set
-p b.eval(["Red", "SUV", "Domestic"])
-p b.get_probability_map(["Red", "SUV", "Domestic"])
+b = Ai4r::Classifiers::NaiveBayes.new
+                                 .set_parameters({ m: 3 })
+                                 .build data_set
+p b.eval(%w[Red SUV Domestic])
+p b.get_probability_map(%w[Red SUV Domestic])

data/examples/classifiers/one_r_example.rb

@@ -0,0 +1,27 @@
+# frozen_string_literal: true
+
+# Author::    Example contributor
+# License::   MPL 1.1
+# Project::   ai4r
+#
+# Simple example showing how to use OneR with numeric attributes.
+
+require_relative '../../lib/ai4r/classifiers/one_r'
+require_relative '../../lib/ai4r/data/data_set'
+
+items = [
+  ['New York', 20, 'M', 'Y'],
+  ['Chicago', 25, 'M', 'Y'],
+  ['New York', 28, 'M', 'Y'],
+  ['New York', 35, 'F', 'N'],
+  ['Chicago', 40, 'F', 'Y'],
+  ['New York', 45, 'F', 'N'],
+  ['Chicago', 55, 'M', 'N']
+]
+labels = %w[city age gender marketing_target]
+
+ds = Ai4r::Data::DataSet.new(data_items: items, data_labels: labels)
+
+classifier = Ai4r::Classifiers::OneR.new.build(ds)
+puts classifier.get_rules
+puts classifier.eval(['Chicago', 55, 'M'])

data/examples/classifiers/parameter_tutorial.rb

@@ -0,0 +1,29 @@
+# frozen_string_literal: true
+
+# This small tutorial shows how changing parameters of ZeroR and OneR
+# affects the generated rules. Run it with `ruby parameter_tutorial.rb`.
+
+require_relative '../../lib/ai4r/classifiers/zero_r'
+require_relative '../../lib/ai4r/classifiers/one_r'
+require_relative '../../lib/ai4r/data/data_set'
+
+# Load the demonstration data set
+file = "#{File.dirname(__FILE__)}/zero_one_r_data.csv"
+set = Ai4r::Data::DataSet.new.load_csv_with_labels file
+
+puts '== ZeroR with default parameters =='
+zero_default = Ai4r::Classifiers::ZeroR.new.build(set)
+puts zero_default.get_rules
+
+puts "\n== ZeroR with :tie_break => :random =="
+zero_rand = Ai4r::Classifiers::ZeroR.new.set_parameters(tie_break: :random).build(set)
+puts zero_rand.get_rules
+
+puts "\n== OneR default behaviour =="
+one_default = Ai4r::Classifiers::OneR.new.build(set)
+puts one_default.get_rules
+
+puts "\n== OneR forcing first attribute and :last tie break =="
+one_custom = Ai4r::Classifiers::OneR.new.set_parameters(selected_attribute: 0,
+                                                        tie_break: :last).build(set)
+puts one_custom.get_rules

data/examples/classifiers/prism_nominal_example.rb

@@ -0,0 +1,15 @@
+# frozen_string_literal: true
+
+require_relative '../../lib/ai4r/classifiers/prism'
+require_relative '../../lib/ai4r/data/data_set'
+
+data_file = "#{File.dirname(__FILE__)}/zero_one_r_data.csv"
+data = Ai4r::Data::DataSet.new.load_csv_with_labels(data_file)
+
+classifier = Ai4r::Classifiers::Prism.new.build(data)
+
+puts 'Discovered rules:'
+puts classifier.get_rules
+puts
+sample = data.data_items.first[0...-1]
+puts "Prediction for #{sample.inspect}: #{classifier.eval(sample)}"

data/examples/classifiers/prism_numeric_example.rb

@@ -0,0 +1,21 @@
+# frozen_string_literal: true
+
+require_relative '../../lib/ai4r/classifiers/prism'
+require_relative '../../lib/ai4r/data/data_set'
+
+items = [
+  [20, 70, 'N'],
+  [25, 80, 'N'],
+  [30, 60, 'Y'],
+  [35, 65, 'Y']
+]
+labels = %w[temperature humidity play]
+
+data = Ai4r::Data::DataSet.new(data_items: items, data_labels: labels)
+
+classifier = Ai4r::Classifiers::Prism.new.build(data)
+
+puts 'Rules:'
+puts classifier.get_rules
+puts
+puts "Prediction for [30, 70]: #{classifier.eval([30, 70])}"

data/examples/classifiers/simple_linear_regression_example.rb

@@ -1,15 +1,18 @@
-require File.dirname(__FILE__) + '/../../lib/ai4r/classifiers/simple_linear_regression'
-require File.dirname(__FILE__) + '/../../lib/ai4r/data/data_set'
-require 'benchmark'
+# frozen_string_literal: true
 
-include Ai4r::Classifiers
-include Ai4r::Data
+require_relative '../../lib/ai4r/classifiers/simple_linear_regression'
+require_relative '../../lib/ai4r/data/data_set'
 
-data_set = DataSet.new
-data_set.parse_csv_with_labels File.dirname(__FILE__) + "/simple_linear_regression_example.csv"
+# Load training data
+file = "#{File.dirname(__FILE__)}/simple_linear_regression_example.csv"
+data_set = Ai4r::Data::DataSet.new.parse_csv_with_labels file
 
-r = SimpleLinearRegression.new.build data_set
-p r.eval([-1,95,109.1,188.8,68.9,55.5,3062,141,3.78,3.15,9.5,114,5400,19,25])
+# Build the regression model and inspect its coefficients
+r = Ai4r::Classifiers::SimpleLinearRegression.new.build data_set
+puts "Selected attribute: #{r.attribute}"
+puts "Slope: #{r.slope}, Intercept: #{r.intercept}"
 
-# => 11662.949367088606
-#Actual price 22625
+# Predict a new sample
+predicted = r.eval([-1, 95, 109.1, 188.8, 68.9, 55.5, 3062, 141, 3.78, 3.15, 9.5, 114, 5400, 19,
+                    25])
+puts "Predicted value: #{predicted}"

data/examples/classifiers/zero_and_one_r_example.rb

@@ -0,0 +1,34 @@
+# frozen_string_literal: true
+
+require_relative '../../lib/ai4r/classifiers/zero_r'
+require_relative '../../lib/ai4r/classifiers/one_r'
+require_relative '../../lib/ai4r/data/data_set'
+
+# Load tutorial data
+data_file = "#{File.dirname(__FILE__)}/zero_one_r_data.csv"
+data = Ai4r::Data::DataSet.new.load_csv_with_labels data_file
+
+puts "Data labels: #{data.data_labels.inspect}"
+puts
+
+# Build a default ZeroR classifier
+zero_default = Ai4r::Classifiers::ZeroR.new.build(data)
+puts "ZeroR default prediction: #{zero_default.eval(data.data_items.first)}"
+puts "Generated rule: #{zero_default.get_rules}"
+
+# Build ZeroR with custom tie strategy
+zero_random = Ai4r::Classifiers::ZeroR.new.set_parameters(tie_break: :random).build(data)
+puts "ZeroR random tie strategy prediction: #{zero_random.eval(data.data_items.first)}"
+
+puts
+
+# Build a default OneR classifier
+one_default = Ai4r::Classifiers::OneR.new.build(data)
+puts "OneR chose attribute index #{one_default.rule[:attr_index]}"
+puts "OneR rules:\n#{one_default.get_rules}"
+
+# Build OneR selecting the first attribute and using :last tie break
+one_custom = Ai4r::Classifiers::OneR.new.set_parameters(selected_attribute: 0,
+                                                        tie_break: :last).build(data)
+puts "OneR forced attribute: #{one_custom.rule[:attr_index]}"
+puts "Custom rules:\n#{one_custom.get_rules}"

data/examples/classifiers/zero_one_r_data.csv

@@ -0,0 +1,8 @@
+city,age_range,gender,marketing_target
+New York,[30-50),F,N
+New York,<30,M,Y
+Chicago,<30,M,Y
+New York,<30,M,Y
+Chicago,[30-50),F,Y
+New York,[30-50),F,N
+Chicago,[50-80],M,N

data/examples/clusterers/clusterer_example.rb

@@ -1,56 +1,62 @@
+# frozen_string_literal: true
+
 # Author::    Sergio Fierens (implementation)
 # License::   MPL 1.1
 # Project::   ai4r
 # Url::       http://www.ai4r.org/
 #
-# You can redistribute it and/or modify it under the terms of 
-# the Mozilla Public License version 1.1  as published by the 
+# You can redistribute it and/or modify it under the terms of
+# the Mozilla Public License version 1.1  as published by the
 # Mozilla Foundation at http://www.mozilla.org/MPL/MPL-1.1.txt
 
 # In this example we group results from a post-training survey into 4 groups.
-# The Diana algorithm is used, but you can try other algorithms by changing 
+# The Diana algorithm is used, but you can try other algorithms by changing
 # the word "Diana" by "KMeans", "AverageLinkage", or any other cluster implementation.
 # The cluster API is the same, so you can play around and observe different results.
 
-require 'rubygems'
 require 'ai4r'
-include Ai4r::Data
-include Ai4r::Clusterers
 
 # 5 Questions on a post training survey
-questions = [	"The material covered was appropriate for someone with my level of knowledge of the subject.", 
-				"The material was presented in a clear and logical fashion", 
-				"There was sufficient time in the session to cover the material that was presented", 
-				"The instructor was respectful of students", 
-				"The instructor provided good examples"]
+questions = ['The material covered was appropriate for someone with my level of ' \
+             'knowledge of the subject.',
+             'The material was presented in a clear and logical fashion',
+             'There was sufficient time in the session to cover the material that was presented',
+             'The instructor was respectful of students',
+             'The instructor provided good examples']
 
 # Answers to each question go from 1 (bad) to 5 (excellent)
-# The answers array has an element per survey complemented. 
+# The answers array has an element per survey complemented.
 # Each survey completed is in turn an array with the answer of each question.
-answers = [	[ 1, 2, 3, 2, 2],	# Answers of person 1
-		[ 5, 5, 3, 2, 2],	# Answers of person 2
-		[ 1, 2, 3, 2, 2],	# Answers of person 3
-		[ 1, 2, 2, 2, 2],	# ...
-		[ 1, 2, 5, 5, 2],
-		[ 3, 3, 3, 3, 3],
-		[ 1, 2, 3, 2, 2],
-		[ 3, 2, 3, 5, 5],
-		[ 3, 3, 3, 5, 2],
-		[ 4, 4, 3, 1, 1],
-		[ 5, 5, 5, 5, 5],
-		[ 4, 2, 4, 2, 1],
-		[ 4, 4, 5, 5, 5],
-		[ 4, 4, 3, 2, 2],
-		[ 2, 2, 3, 2, 3],
-		[ 3, 3, 3, 1, 1]]	# Answers of person 16
-
-data_set = DataSet.new(:data_items => answers, :data_labels => questions)
+answers = [[1, 2, 3, 2, 2],	# Answers of person 1
+           [5, 5, 3, 2, 2],	# Answers of person 2
+           [1, 2, 3, 2, 2],	# Answers of person 3
+           [1, 2, 2, 2, 2],	# ...
+           [1, 2, 5, 5, 2],
+           [3, 3, 3, 3, 3],
+           [1, 2, 3, 2, 2],
+           [3, 2, 3, 5, 5],
+           [3, 3, 3, 5, 2],
+           [4, 4, 3, 1, 1],
+           [5, 5, 5, 5, 5],
+           [4, 2, 4, 2, 1],
+           [4, 4, 5, 5, 5],
+           [4, 4, 3, 2, 2],
+           [2, 2, 3, 2, 3],
+           [3, 3, 3, 1, 1]]	# Answers of person 16
+
+data_set = Ai4r::Data::DataSet.new(data_items: answers, data_labels: questions)
 
 # Let's group answers in 4 groups
-clusterer = Diana.new.build(data_set, 4)
+clusterer = Ai4r::Clusterers::Diana.new.build(data_set, 4)
 
-clusterer.clusters.each_with_index do |cluster, index| 
-	puts "Group #{index+1}"
-	p cluster.data_items
+clusterer.clusters.each_with_index do |cluster, index|
+  puts "Group #{index + 1}"
+  p cluster.data_items
 end
 
+# Check if this algorithm supports evaluating new data items
+if clusterer.supports_eval?
+  puts "First survey belongs to group #{clusterer.eval(answers.first)}"
+else
+  puts 'This algorithm does not support eval on unseen data.'
+end

data/examples/clusterers/dbscan_example.rb

@@ -0,0 +1,17 @@
+# frozen_string_literal: true
+
+# Example showcasing DBSCAN clustering with custom parameters.
+require 'ai4r'
+
+points = [
+  [1, 1], [1, 2], [1, 3], [2, 1], [2, 2], [2, 3],
+  [8, 8], [8, 9], [8, 10], [9, 8], [9, 9], [9, 10],
+  [5, 5], [1, 9], [10, 0]
+]
+set = Ai4r::Data::DataSet.new(data_items: points)
+
+clusterer = Ai4r::Clusterers::DBSCAN.new
+clusterer.set_parameters(epsilon: 10, min_points: 2).build(set)
+
+pp clusterer.labels
+pp clusterer.clusters.map(&:data_items)

data/examples/clusterers/dendrogram_example.rb

@@ -0,0 +1,17 @@
+# frozen_string_literal: true
+
+require 'ai4r'
+require 'dendrograms'
+
+points = [[0, 0], [0, 1], [1, 0], [1, 1]]
+data = Ai4r::Data::DataSet.new(data_items: points)
+
+clusterer = Ai4r::Clusterers::WardLinkage.new.build(data, 1)
+
+# Convert stored tree to a simple array of point sets
+steps = clusterer.cluster_tree.map do |clusters|
+  clusters.map(&:data_items)
+end
+
+Dendrograms::Dendrogram.new(steps).draw('dendrogram.png')
+puts 'Dendrogram saved to dendrogram.png'