AI Primer for Marketers Part 2: Algorithms and Machine Learning Basics

营销策划 2017-06-06

The most innovative marketers routinely pick up new things, try them out, and succeed or fail. Why are marketers struggling so much to adapt to artificial intelligence and machine learning? In this series, we’ll explore machine learning and artificial intelligence to build a foundation for understanding the field – and how it applies to marketing.

What is Artificial Intelligence?

Before we can begin discussing how artificial intelligence and machine learning will impact marketing, we have to establish some basic definitions.

First, artificial intelligence is the science of creating computer hardware and software that mimics human intelligence and performs human intelligence functions. For example, if you are reading the words on the screen right now and they don’t appear as gibberish, you are using visual recognition and natural language processing, two fields of study in artificial intelligence.

The Basics of AI: Algorithms

The foundation of artificial intelligence, and of computing itself, begins in the algorithm. Named after a 9th century Persian mathematician, an algorithm is a set of repeatable processes that deliver a reliable, repeatable result. We use algorithms every day. Our morning routine is an algorithm. The way we make our coffee is an algorithm. The way we drive to work is a complex series of algorithms.

In marketing, we’ve been using and discussing algorithms since the dawn of digital marketing. Our first experience with algorithms was in SEO, as we tried to figure out what pages ranked well insearch engines. We developed our own algorithms for creating content to be found by search engines. Social media marketing is entirely based in algorithms. We talk about them everyday – the Facebook news feed algorithm, the Instagram algorithm, how these platforms choose what content appears to users.

In the basics of computing, algorithms don’t change by themselves. We have to create them and modify them every time something changes. For example, if we’re out of sweetener, we have to change the way we make our coffee. If there’s a traffic jam, we have to change how we drive to work.

What if algorithms could change themselves based on new information?

They can – and that’s what we call machine learning.

The Basics of AI: Machine Learning

Machine learning is exactly as it sounds: the ability for machines to learn without being explicitly programmed. Given new data, a machine can adjust its own algorithms to be more efficient or more effective.

For example, we use machine learning every time we use our smartphone GPS. We put in our destination and the GPS finds the most efficient route for us. If traffic conditions change, our GPS changes along with those conditions and finds us a new way to get to where we’re going.

As consumers, we’ve had experience with machine learning since the early days of a digital marketing. Every time a platform or a service remembers our preferences and changes with them, we are seeing machine learning at work. Every time we shop online and a website gives us a list of recommendations that get better and better the more it gets to know us, we are seeing machine learning at work.

The foundations of machine learning are in algorithms and statistics. For example, when we are shopping online and a website is deciding what to show us as a recommended item to go along with our purchase, the machine learning algorithm is running a series of statistical tests.

Based on past experience and what other customers bought, what is the probability that we will like and purchase the red item versus the blue item? each time we purchase something, the website learns and scores its previous statistical test.

If it succeeded in convincing us to buy something additional in our order, then it updates as probability calculations for the next customer and for the next time we come back. If it didn’t convince us to buy something, then it will rerun its statistical tests to find what else we might purchase instead.

Next: Types of Machine Learning

In the next post in this series, we will look at the different types of machine learning and how they apply to marketing. Stay tuned!

If you enjoyed this, please share it with your network!

Want to read more like this from Christopher Penn ? Get updates here:

Christopher S. Penn

责编内容by:Christopher S. Penn (源链)。感谢您的支持!


HAWQ + MADlib 玩转数据挖掘之(十二)——模型评估之交叉验证... 一、交叉验证概述 机器学习技术在应用之前使用“训练+检验”的模式,通常被称作“交叉验证”,如图1所示。 图1 1. 预测模型的稳定性 ...
独家揭秘:腾讯千亿级参数分布式ML系统无量背后的秘密... 作者 | 袁镱 编辑 | Vincent AI 前线导读: 千亿参数规模的模型已经被业界证明能够有效提高业务效果。如何高效训练出这样的...
Box deepens partnership with Microsoft and turns i... When I spoke to Box CEO Aaron Levie last year at the Boxworks customer co...
个性化推荐是怎么做的?产品经理也可以懂的算法... 今日头条带动了“个性化推荐”的概念,自此之后,无论是工具产品,电商产品,还是内容型的产品,都自带内容属性,个性化算法也逐渐从卖点变为标配。 各种推荐算法不能...
使用 Python 实现接缝裁剪算法 接缝裁剪是一种新型的裁剪图像的方式,它不会丢失图像中的重要内容。这通常被称之为“内容感知”裁剪或图像重定向。你可以从这张照片中感受一下这个算法: ...