NDC Oslo : A Practical Introduction to Data Science

A Practical Introduction to Data Science @markawest

A Practical-ish Introduction to Data Science @markawest

Who Am I? • Previously Java Developer and Architect. @markawest

Who Am I? • Previously Java Developer and Architect. • Currently building and managing a team of Data Scientists at Bouvet Oslo. @markawest

Who Am I? • Previously Java Developer and Architect. • Currently building and managing a team of Data Scientists at Bouvet Oslo. • Leader javaBin (Norwegian Java User Group). @markawest

Agenda What is Data Science? Machine Learning Algorithms Practical Example @markawest

What is Data Science? What is Data Science? Machine Learning Algorithms Practical Example @markawest

@markawest “An estimated 2.5 quintillion* bytes of data are created each day…” IBM * 1 quintillion = 1,000,000,000,000,000,000 (or 1,000,000,000 Gigabytes)

@markawest “Data Science… is an interdisciplinary field of scientific methods, processes, and systems to extract knowledge or insight from data…” Wikipedia

Computer Science/IT @markawest

Computer Science/IT Domain/Business Knowledge Software Development @markawest

Computer Science/IT Math and Statistics Domain/Business Knowledge Machine Learning Software Development Traditional Research Data Science @markawest

Computer Science/IT Math and Statistics Domain/Business Knowledge Machine Learning Software Development Traditional Research @markawest

@markawest 1. Question 2. Data 3. Exploratory Data Analysis 4. Formal Modelling 5. Interperetation 6. Communication 7. Result Data Science Process : Hypothesis Driven

@markawest Roles Required in a Data Science Project • Prove / disprove hypotheses. • Information and Data Gathering. • Data Wrangling. • Algorithm and ML models. • Communication. Data Scientist • Build Data Driven Platforms. • Operationalize Algorithms and Machine Learning models. • Data Integration. Data Engineer • Storytelling. • Build Dashboards and other Data visualizations. • Provide insight through visual means. Visualization Expert • Project Management. • Manage stakeholder expectations. • Maintain a Vision. • Facilitate. Process Owner

@markawest Roles Required in a Data Science Project • Prove / disprove hypotheses. • Information and Data gathering. • Data wrangling. • Algorithm and ML models. • Communication. Data Scientist • Build Data Driven Platforms. • Operationalize Algorithms and Machine Learning models. • Data Integration. Data Engineer • Storytelling. • Build Dashboards and other Data visualizations. • Provide insight through visual means. Visualization Expert • Project Management. • Manage stakeholder expectations. • Maintain a Vision. • Facilitate. Process Owner

Isn’t Data Science just a rebranding of Business Intelligence? @markawest

@markawest Data Science as an Evolution of BI Business Intelligence Data Science Adds.. Data Sources Structured Data, most often from Relational Database Management Systems (RDBMS). Unstructured Data (log files, audio, images, emails, tweets, raw text, documents). Available Tools Data Visualization, Statistics. Advanced technologies such as Machine Learning and NLP. Goals Provide support to strategic decision making, based on historical data. Provide business value through advanced functionality and technologies. Source: https://www.linkedin.com/pulse/data-science-business-intelligence-whats-difference-david-rostcheck

@markawest Machine Learning: A Tool for Data Science

@markawest Machine Learning: A Tool for Data Science Artificial Intelligence Artificial Intelligence Enabling computers to mimic human intelligence and behavior.

@markawest Machine Learning: A Tool for Data Science Artificial Intelligence Machine Learning Artificial Intelligence Enabling computers to mimic human intelligence and behavior. Machine Learning Algorithms allowing computers to learn, make predictions and describe data without being explicitly programmed.

@markawest Machine Learning: A Tool for Data Science Artificial Intelligence Machine Learning Deep Learning Machine Learning Algorithms allowing computers to learn, make predictions and describe data without being explicitly programmed. Artificial Intelligence Enabling computers to mimic human intelligence and behavior. Deep Learning Black box learning with multi-layered Neural Networks.

What is Data Science: Key Takeaways • Data Scientists require Math and Statistics skills in addition to traditional Software Development. • Data Science is Hypothesis Driven. • Data Science projects require a range of competencies/roles. • Data Science can be seen as an evolution of Business Intelligence, providing additional capabilities through the application of cutting edge technologies and unstructured data. @markawest

Machine Learning Algorithms What is Data Science? Machine Learning Algorithms Practical Example @markawest

@markawest “Machine Learning: Field of study that gives computers the ability to learn without being explicitly programmed.” Arthur L. Samuel IBM Journal of Research and Development, 1959 Computer Data Rules Output Computer Data Output Rules Traditional Programming Machine Learning

Machine Learning: Models vs. Algorithms @markawest Algorithm Training Data (In) Model (Out) Data set representing the domain to be modelled. Implemented Machine Learning Algorithm. Rule set based on Algorithm and Training Data.

The Importance of a Generalized Model @markawest Training Data New Data

Overfitting & Underfitting in Machine Learning @markawest Underfitted Appropriate Overfitted Generalized model with an acceptable error margin. Model focuses on noise in training data. Model overlooks underlying patterns.

Supervised Learning Machine Learning Types @markawest Unsupervised Learning Model trained on historical data. Resulting model can be used to make predictions on new data. Use Case: Predicting a value based on patterns discovered in previous data. Model finds trends and patterns in data, without prior training on historical data. Use Case: Describing your data based on statistical analysis. Reinforcement Learning Model uses a feedback loop to iteratively improve it’s performance. Use Case: Learning how to best solve a problem based on trial and error.

Machine Learning Algorithm Types @markawest Supervised Learning Unsupervised Learning

Common Machine Learning Types @markawest Supervised Learning Unsupervised Learning ClassificationRegression Clustering

Example Machine Learning Algorithms @markawest Supervised Learning Unsupervised Learning Linear Regression ClassificationRegression K-Means Clustering Support Vector Machines Decision Trees

Floor Space House Price 1 180 221 900 2 570 538 000 770 180 000 1 960 604 000 1 680 510 000 … … … … 5 240 1 225 000 Linear Regression (predicts continuous values) Feature Label @markawest

Floor Space House Price 1 180 221 900 2 570 538 000 770 180 000 1 960 604 000 1 680 510 000 … … … … 5 240 1 225 000 Linear Regression (predicts continuous values) Feature Label Trend Line Deviation Outlier Prediction @markawest

Linear Regression Notes Benefits • Simple to understand. • Transparent. Limitations • Outliers skew trend line. • Doesn’t work well with non- linear relationships. Some Alternatives • Non-linear Least Squares. • Tree algorithms. @markawest

Support Vector Machine (predicts discrete classes) @markawest • Challenge: Define an optimal line for separating classes.

Support Vector Machine (predicts discrete classes) @markawest • Challenge: Define an optimal line for separating classes. • Solution: Find the support vectors, the optimal boundary can be placed equidistant between these.

Support Vector Machine (predicts discrete classes) @markawest • Challenge: Define an optimal line for separating classes. • Solution: Find the support vectors, the optimal boundary can be placed equidistant between these. Optimal Boundary Support Vector Support Vector

The SVM Buffer Zone (predicts discrete classes) @markawest • SVM’s allow the definition of a “buffer zone”. • Any training data in the buffer zone is ignored, making the model resilient to outliers and more generic. Optimal Boundary Support Vector Support Vector

Support Vector Machine Notes Benefits • Resistant to outliers. • Works with both linear and non-linear boundaries. • Works well with large feature sets (high dimensionality). Limitations • Works best with binary classification. • Tricky to tune (i.e. buffer zone). Some Alternatives • Logistic Regression. • Decision Trees. • K-Nearest Neighbors. @markawest

Decision Tree (predicts discrete classes) @markawest Outlook Temp Humidity Wind Play Sunny Hot High Weak No Sunny Hot High Strong No Overcast Hot High Weak Yes … … … … … … … … … … Overcast Mild High Strong Yes Overcast Hot Normal Weak Yes Rain Mild High Strong No No Yes No Yes Yes Outlook Humidity Wind Features Labels Overcast Sunny Rain High WeakNormal Strong

Building a Decision Tree: Recursive Partitioning @markawest 1. Find the Feature that optimally splits data points into homogeneous groups.Outlook Temp Humidity Wind Play Sunny Hot High Weak No Sunny Hot High Strong No Overcast Hot High Weak Yes … … … … … … … … … … Overcast Mild High Strong Yes Overcast Hot Normal Weak Yes Rain Mild High Strong No Features Labels

Building a Decision Tree: Recursive Partitioning @markawest 1. Find the Feature that optimally “slices” data points into homogeneous groups. 2. Repeat for each leaf, until a stopping criteria is reached, for example: • Data points at each leaf have the same value. • Another threshold is set (i.e. maximum tree depth or minimum data points for each leaf is reached). No Yes No Yes YesHumidity Wind Overcast Sunny Rain High Normal Strong Outlook

Decision Tree Notes Benefits • White Box. • Flexible (use for both regression and classification). • Robust to outliers. • Handle non-linear boundaries. Limitations • Susceptible to overfitting. • Changes to where the Data is sliced can produce different results. Some Alternatives • Support Vector Machine. • Logistic Regression. • Random Forests. @markawest

K-Means Clustering (describing data without training first) @markawest • K = The amount of clusters the algorithm will try to find. • K = Should be large enough to extract meaningful patterns but small enough that clusters remain clearly distinct. • So how do we calculate K?

K-Means: Calculating the K value @markawest • Scree Plots allow us to find optimal number of clusters. • Sum of Squared Error (SSE). • The optimal K value is at the “Elbow” of the plot.

K-Means Demo Randomly allocate centroids @markawest

K-Means Demo Iteration 1: Calculate cluster membership based on nearest centroid @markawest

K-Means Demo Iteration 1: Move centroids to the center of their cluster @markawest

K-Means Demo Iteration 2: Recalculate cluster membership based on nearest centroid @markawest

K-Means Demo Iteration 2: Move centroids to the center of their cluster @markawest

K-Means Demo After 6 iterations: Clusters and centroids stablise, algorithm stops @markawest

K-Means Clustering Notes Benefits • Fast and highly effective at uncovering basic data patterns. • Works best for spherical, non- overlapping clusters. Limitations • Each data point can only be assigned to one cluster. • Clusters are assumed to be spherical. Some Alternatives • Gaussian mixtures. • Fuzzy K-Means. @markawest

Does Machine Learning really need Data Science skillz? @markawest

Data Science Skills and Machine Learning @markawest Machine Learning Data In Data Out Variable Selection Feature Engineering Algorithm Selection Algorithm Tuning Interpret Results Evaluation Communication

Machine Learning Algorithms: Key Takeaways @markawest • The three main types of Machine Learning are Supervised, Unsupervised and Reinforcement Learning. • A successful Machine Learning Model needs to find the balance between Overfitting and Underfitting. • Machine Learning Algorithms are merely tools. Good results come from understanding how they work and tuning them correctly. • Data Science skills are vital for succeeding with Machine Learning.

Practical Example What is Data Science? Machine Learning Algorithms Practical Example @markawest

Use Case: Titanic Passenger Survival @markawest Goal: Build a classification model for predicting Titanic survivability.

Hypothesis That it is possible to predict Titanic survivability based on Age, Gender and Ticket Class. @markawest

@markawest Variable Description PassengerId Unique Identifier Survival Survived = 1, Died = 0 Pclass Ticket class (1, 2 or 3) Sex Gender (‘male’ or ’female’) Age Age in years Sibsp Number siblings / spouses aboard the Titanic Parch Number parents / children aboard the Titanic Ticket Ticket number Fare Passenger fare Cabin Cabin number Embarked Port of Embarkation Name Passenger name, including honorific. Titanic Dataset

Practical Example: Key Takeaways @markawest • Scikit-learn and Jupyter Notebooks provide a free and flexible basis for starting with Data Science. • Feature Engineering is a vital skill for Data Scientists. • Domain Knowledge is key! • Split your data into Test and Training sets. • Tweaking your ML Algorithm Hyperparameters can give better results.

Thanks for listening! @markawest

NDC Oslo : A Practical Introduction to Data Science

More Related Content

What's hot

Similar to NDC Oslo : A Practical Introduction to Data Science

More from Mark West

Recently uploaded

NDC Oslo : A Practical Introduction to Data Science

Editor's Notes