Data science, machine learning, and analytics without coding

Name: Data science, machine learning, and analytics without coding
Rating: 4.3 (50 reviews)

Solve real data science problems and add value quickly without needing to learn how to code

Created byEric Hulbert

Last updated 1/2021

English

What you'll learn

The fundamentals of data science problem solving
Machine learning algorithms such as Random Forest, K-Means, and OLS Regression
How to use the KNIME platform to import, process, explore, and clean data

Course content

8 sections • 44 lectures • 3h 10m total length

Intro: why this course is the best option for those wanting to make an impact2:48
About your instructor1:40
Eric champions no-code data exploration and cleaning using tools like Altair six and nine, guiding clients from problem framing to turning data insights into solutions.
What is KNIME, the platform we will use in this course?3:28
How to get KNIME (don't worry, it is free!) - do this before we start0:26

Intro to data exploration3:44
Use exploratory data analysis to gain data understanding, check data quantity and labeling, and identify quality issues that can derail modeling.
KNIME tour and importing / accessing data5:51
Data types in KNIME and data analysis everywhere2:58
Explore the five core data types—double (and float), integer, string, date/time, and boolean—in KNIME and any data analysis tool, with notes on math, comparisons, and joins.
Group By: The most powerful node in KNIME4:23
Pivoting: Using pivots and the math formula for more value out of Group By6:04
Statistics: Using summary statistics from the nodes2:25
Explore summary statistics from the statistics node, including mean, standard deviation, variance, skew, and a histogram to reveal data distribution across three shifts and guide data cleaning.
Visualization: Graphing & plotting in KNIME2:37

Intro to data cleaning3:50
Learn data cleaning through exploration, iterative fixes of strings and duplicates, and handling missing data and outliers to prepare data for modeling.
String to Date: Changing strings to dates5:15
Convert a date-like string to a date type using the string to date step in time series. Use the format d-MM-yy, noting capital M denotes month and the dash matters.
String Manipulation: Fixing our strings so we can use them8:06
Learn practical string manipulation to clean data: convert numbers to strings, remove periods and commas, strip exponential notation, and pad identifiers to a fixed length.
Metanodes: Combine nodes to clean up your workflow2:31
Learn to clean up your data workflow by combining nodes into metanodes, label steps, and fit large programs into a single node you can open to inspect.
String Manipulation (part deux): So important we do it twice4:09
Rule engine: How to use "if-else" statements in KNIME3:23
Row Filter: Removing rows we do not want to analyze3:00
Apply row filtering to remove unwanted data by color-based and rule-based criteria, using pattern matching and membership checks to exclude red and black entries and keep orange and blue.
Missing Data: How KNIME can help us deal2:41

Call Center Collections (CCC) business and data situation4:16
Linear regression and how to perform it correctly5:14
Implementing linear regression part 1: Building9:01
Learn to implement a linear regression model, diagnose issues using a correlation matrix, and select predictors by removing highly correlated features to predict cash, with 80/20 partition and R-squared.
Implementing linear regression part 2: Refining3:22
Implementing linear regression part 3: Checking assumptions5:28
Evaluating linear regression3:27
What can we do with this?3:05

Bob's Best Boats - business and data understanding3:10
K-Means clustering - how it works5:49
Performing our first K-Means cluster3:53
Correcting our first K-Means cluster2:33
Normalize data with a z-score before applying k-means to ensure Euclidean distance reflects the income axis, highlighting why scaling incomes and credit scores matters for clustering.
Evaluating the clusters - are they any good?6:57
Creating better segments for the client9:10
Explore data-driven segmentation with decision trees and pattern aggregation to reveal segments like divorced customers with kids and income-based bins for targeted boat rental marketing.

Requirements

A computer with enough space to install the KNIME Analytics Platform

Description

Do you want to super charge your career by learning the most in demand skills? Are you interested in data science but intimidated from learning by the need to learn a programming language?

I can teach you how to solve real data science business problems that clients have paid hundreds of thousands of dollars to solve. I'm not going to turn you into a data scientist; no 2 hour, or even 40 hour online course is able to do that. But this course can teach you skills that you can use to add value and solve business problems from day 1.

This course is different than most for several reasons:

1. We start with problem solving instead of coding. I feel like starting to code before solving problems is misguided; many students are turned off by hours of work to try to write a couple of meaningless lines rather than solving real problems. The key value add data scientists make is solving problems, not writing something in a language a computer understands.

2. The examples are based on real client work. This is not like other classes that use Kaggle data sets for who survived the Titanic, or guessing what type of flower it is based on petal measurements. Those are interesting, but not useful for people wanting to sell more products, or optimize the performance of their teams. These examples are based on real client problems that companies spent big money to hire consultants (me) to solve.

3. Visual workflows. KNIME uses a visual workflow similar to what you'll see in Alteryx or Azure Machine Learning Studio and I genuinely think it is the future of data science. It is a better way of visualizing the problem as your are exploring data, cleaning data, and ultimately modeling. It is also something that makes your process far easier to explain to non-data scientists making it easier to work with other parts of your business.

Summary: This course covers the full gamut of the machine learning workflow, from data and business understanding, through exploration, cleaning, modeling, and ultimately evaluation of the model. We then discuss the practical aspects of what you can change, and how you can change it, to drive impact in the business.

Who this course is for:

Beginners in data science who do not know how to code
People who want to learn data science problem solving but do not think they will be able to learn code
Business people who want to solve problems that are too large or difficult for Excel

Data science, machine learning, and analytics without coding

What you'll learn

Explore related topics

Course content

Introduction4 lectures • 8min

Data Exploration7 lectures • 28min

Data Cleaning8 lectures • 33min

Modeling Overview3 lectures • 10min

Client model #1: Random Forest for predicting sales outcomes7 lectures • 43min

Client model #2: Linear regression for call center performance improvement7 lectures • 34min

Client model #3 - K-Means and clustering to find attractive segments6 lectures • 32min

If you are interested in learning more2 lectures • 3min

Requirements

Description

Who this course is for: