Machine learning and deep learning have become an important part of many of the applications we use every day. There are few domains whose rapid expansion is machine learning has not touched. Many companies have developed the right strategy to incorporate machine learning algorithms into their operations and processes. Others have lost ground to their competitors after ignoring undeniable advances in artificial intelligence.
However, mastering machine learning is a difficult process. You need to start with a solid understanding of linear algebra and calculus, master a programming language like Python, and master data science and machine learning libraries like Numpy, Scikit-learn, TensorFlow, and more PyTorch.
If you want to build machine learning systems that integrate and scale, you need to learn about cloud platforms like Amazon AWS, Microsoft Azure, and Google Cloud.
Of course, not everyone has to become a machine learning engineer. But almost anyone who runs a company or organization that systematically collects and processes can benefit from some knowledge of data science and machine learning. Fortunately, there are several courses out there that provide a comprehensive overview of machine learning and deep learning without getting too into math and coding.
In my experience, a good understanding of data science and machine learning requires some hands-on experience with algorithms. In this regard, Microsoft Excel is a very valuable and often overlooked tool.
“Learn Data Mining Using Excel: A Step-by-Step Approach to Understanding Machine Learning Methods” by Hong Zhou
For most people, MS Excel is a spreadsheet application that stores data in tabular form and performs very basic math operations. In reality, however, Excel is a powerful calculation tool that can be used to solve complicated problems. Excel also has many features that you can use to build machine learning models right in your workbooks.
While using Excel’s math tools for years, I didn’t really appreciate its use to learn and apply data science and machine learning until I started Learn Data Mining Via Excel: A Step-by-Step Approach to Understanding Machine Learning Methods by Hong Zhou.
Learning data mining through Excel takes you step-by-step through the basics of machine learning and shows how you can implement many algorithms using basic Excel functions and some of the application’s advanced tools.
While it will in no way replace Excel Python Machine Learning, it’s a great window to learn the basics of AI and solve a lot of basic problems without writing a line of code.
Machine learning with linear regression with Excel
Linear regression is a simple machine learning algorithm that has many uses for analyzing data and predicting outcomes. Linear regression is particularly useful when your data is organized in a neat tabular format. Excel has several features that you can use to build regression models from tabular data in your tables.
One of the most intuitive is the Data Chart Tool, a powerful data visualization feature. For example, the scatter plot shows the values of your data on a Cartesian plane. In addition to showing the distribution of your data, Excel’s charting tool can create a machine learning model that can predict the changes in the values of your data. The function called Trendline creates a regression model from your data. You can set the trendline to any of several regression algorithms, including linear, polynomial, logarithmic, and exponential. You can also configure the graph to show the parameters of your machine learning model that you can use to predict the outcome of new observations.
You can add multiple trend lines to the same chart. This allows you to quickly test and compare the performance of different machine learning models based on your data.
With the Trendline function of Excel, regression models can be created from your data.
In addition to exploring the chart tool Learning data mining through Excel takes you through various other techniques that you can use to develop advanced regression models. This includes formulas such as LINEST and LINREG formulas that compute the parameters of your machine learning models based on your training data.[Read: How Netflix shapes mainstream culture, explained by data]
The author also walks you through the step-by-step process of building linear regression models using the basic Excel formulas such as SUM and SUMPRODUCT. This is a recurring topic in the book: you see the mathematical formula of a machine learning model, learn the basic arguments behind it, and create it step by step by combining values and formulas in multiple cells and cell arrays.
This may not be the most efficient way to do data science at the production level, but it is certainly a very good way to learn how machine learning algorithms work.
Other machine learning algorithms using Excel
In addition to regression models, you can use Excel for other machine learning algorithms. Learn Data Mining via Excel offers an extensive list of supervised and unsupervised machine learning algorithms, including k-means clustering, k-nearest neighbor, naive Bayesian classification, and decision trees.
The process can be a little complicated at times, but if you stay on the right track the logic will easily come into effect. For example, in the k-means clustering chapter, you can use a variety of Excel formulas and functions (INDEX, IF, AVERAGEIF, ADDRESS, and many others) in multiple worksheets to calculate and refine cluster centers. This is not a very efficient method of clustering. You can track and examine your clusters as they are refined on each successive leaf. From an educational perspective, the experience is very different from programming books in which you provide a machine learning library that supports your data points and outputs the clusters and their properties.
When you do k-means clustering in Excel, you can see the refinement of your clusters on consecutive sheets.
In the chapter on the decision tree, you will walk through the process of calculating entropy and selecting functions for each branch of your machine learning model. Again, the process is slow and manual, but going under the hood of the machine learning algorithm is a rewarding experience.
In many of the chapters in the book you use the solver tool to minimize your loss function. This is where you can see the limitations of Excel, as even a simple model with a dozen parameters can slow your computer down to the point of crawling, especially if your data pattern is several hundred rows in size. However, the solver is a particularly powerful tool if you want to optimize the parameters of your machine learning model.
Excel’s solver tool optimizes the parameters of your model and minimizes loss functions
Deep learning and natural language processing with Excel
Learning data mining through Excel shows that Excel can even refine machine learning algorithms. There is a chapter devoted to the careful creation of Deep learning models. First, you’ll create a single layer artificial neural network with less than a dozen parameters. Then you expand the concept to create a deep learning model with hidden layers. The calculation is very slow and inefficient, but it works and the components are the same: cell values, formulas and the powerful solver tool.
Deep learning with Microsoft Excel gives you an insight into how deep neural networks work.
In the last chapter you create a rudimentary one Natural language processing (NLP) application that uses Excel to build a machine learning model for sentiment analysis. With the help of formulas, you can create a “bag of words” model, preprocess hotel ratings, tokenize and classify them based on the density of positive and negative keywords. You will learn a lot about how contemporary AI deals with language and language how much different It’s the way we humans process written and spoken language.
Excel as a machine learning tool
Whether you’re making C-level decisions in your company, working in HR, or managing supply chains and manufacturing facilities, a basic understanding of machine learning is essential when working with data scientists and AI staff. Likewise, if you’re a reporter covering AI news or a PR agency working for a company that uses machine learning and writing about the technology not knowing how it works is a bad idea (I’ll write a separate post about the many awful AI pitches I get every day). In my opinion, Learn data mining through Excel is a smooth and quick read to help you gain this important knowledge.
In addition to learning the basics, Excel can be a powerful addition to your repertoire of machine learning tools. While it is not good for dealing with large amounts of data and complicated algorithms, it can be useful for visualizing and analyzing smaller stacks of data. The results you get with a quick Excel mining operation can provide relevant insight into choosing the right direction and machine learning algorithm to address the problem at hand.
This article was originally published by Ben Dickson on TechTalks, a publication that examines technology trends, how they affect the way we live and do business, and what problems they solve. But we also discuss the evil side of technology, the darker effects of the new technology, and what to look out for. You can read the original article here.