## Machine Learning 101 – Clustering, Regression and Classification

In my last post of this series, I explained the concept of supervised, unsupervised and semi-supervised machine learning. In this post, we will go a bit deeper into machine learning (but don’t worry, it won’t be that deep yet!) and look at more concrete topics. But first of all, we have to define some terms, which basically derive from statistics or mathematics. These are:

• Features
• Labels

Features are known values, which are often used to calculate results. This are the variables that have an impact on a prediction. If we talk about manufacturing, we might want to reduce junk in our production line. Known features from a machine could then be: Temperature, Humidity, Operator, Time since last service. Based on these Features, we can later calculate the quality of the machine output

Labels are the values we want to build the prediction on. In training data, labels are mostly known, but for the prediction they are not known. When we focus on the machine data example from above, a label would be the quality. So all of the features together make up for a good or bad quality and algorithms can now calculate the quality based on that.

Let’s now go on another “classification” of machine learning techniques. We “cluster” them by supervised/unsupervised.

The first one is clustering. Clustering is an unsupervised technique. With clustering, the algorithm tries to find a pattern in data sets without labels associated with it. This could be a clustering of buying behaviour of customers. Features for this would be the household income, age, … and clusters of different consumers could then be built.

The next one is classification. In contrast to clustering, classification is a supervised technique. Classification algorithms look at existing data and predicts what a new data belongs to. Classification is used for spam for years now and these algorithms are more or less mature in classifying something as spam or not. With machine data, it could be used to predict a material quality by several known parameters (e.g. humidity, strength, color, … ). The output of the material prediction would then be the quality type (either “good” or “bad” or a number in a defined space like 1-10). Another well known sample is if someone would survive the titanic – classification is done by “true” or “false” and input parameters are “age”, “sex”, “class”. If you would be 55, male and in 3rd class, chances are low, but if you are 12, female and in first class, chances are rather high.

The last technique for this post is regression. Regression is often confused with clustering, but it is still different from it. With a regression, no classified labels (such as good or bad, spam or not spam, …) are predicted. Instead, regression outputs continuous, often unbound, numbers. This makes it useful for financial predictions and alike. A common known sample is the prediciton of housing prices, where several values (FEATURES!) are known, such as distance to specific landmarks, plot size,… The algorithms could then predict a price for your house and the amount you can sell it for.

In my next post, I will talk about different algorithms that can be used for such problems.

## Machine Learning 101 – Supervised and Unsupervised Learning

I teach Big Data & Data Science at several universities and I work in that field also. Since I wrote a lot here on Big Data itselve and there are now many young professionals deciding if they want to go for data science, I decided to write a short intro series to machine learning. After this intro, you should be capable of getting deeper into this topic and know where to start. To kick off the series, we’ll go over some basics of machine learning.

One of the main ideas behind that is to find patterns in data and make predictions on that data without the need to develop each and every use-case from scratch. Therefore, a certain number of algorithms are available. These algorithms can be “classified” by how they work. the main two principles (which then can also be spilt) are:

• Supervised Learning
• Unsupervised Learning
• Semi-supervised Learning

With supervised learning, the algorithm learns basically by existing data and learning “from the past”. This means that there is basically a lot of learning data that allows the algorithm to find the patterns by learning from this data. This is also often called “a teacher”. It works closely to how we as humans learn: we get information from our parents, teachers and friends and combine this to make future predictions. Examples are:

• Manufacturing: if several properties of a material were of specific properties, the quality was either good or bad (or maybe scaled from several numbers). Now, if we produce a new material and we look at the properties of the material, based on the existing data we have from former productions, we can say how the quality will be. Properties of a material might be: hardness, color, …
• Banking: based on several properties of a potential borrower, we can predict if the person is capable of paying back the loan. This can be based on existing data of former customers and what the bank “learned” from them. A lot of different variables are calculated for that: income, montly liability to pay, education, job, …

With unsupervised learning we have no “teacher” available. The algorithms get data, and the algorithms try to find patterns in that. This can either be by clustering data (e.g. customer with high income, customer with low income, …) and make predictions based on that. If we look at our industries, this can be used for that:

• Manufacturing: find anomalies in the production lines (e.g. the average output of units per hour was between 200 and 250, but on day D at time T, the output was only 20 units. The algorithm can cluster this into normal output and an anomaly that was detected.
• Banking: normally, the customer would only spend money in his home country. Suddenly, he had high money transfers in a country that he normally isn’t in -> possibility of fraud.

Last, but not least, there is Semi supervised learning, which is a combination of both. In many machine learning projects, not all training data that is used for supervised learning is available, so values might need to get predicted. This can be done by combining supervised and unsupervised learning algorithms and then work with the “curated” data on it.

Now that we basically understand the 3 main concepts, we can continue with variations within these concepts and some statistical background in the next post.

## How to: Start and Stop Cloudera on Azure with the Azure CLI

The Azure CLI is my favorite tool to manage Hadoop Clusters on Azure. Why? Because I can use the tools I am used to from Linux now from my Windows PC. In Windows 10, I am using the Ubuntu Bash for that, which gives me all the major tools for managing remote Hadoop Clusters.

One thing I am doing frequently, is starting and stopping Hadoop Clusters based on Cloudera. If you are coming from Powershell, this might be rather painfull for you, since you can only start each vm in the cluster sequentially, meaning that a cluster consisting of 10 or more nodes is rather slow to start and might take hours! In the Azure CLI I can easily do this by specifiying “–nowait” and all runs in parallel. The only disadvantage is that I won’t get any notifications on when the cluster is ready. But I am doing this with a simple hack: ssh’ing into the cluster (since I have to do this anyway). SSH will succeed once the Masternodes are ready and so I can perform some tasks on the nodes (such as restarting Cloudera Manager since CM is usually a bit “dizzy” after sending it to sleep and waking it up again :))

Let’s start with the easiest step: stopping the cluster. The Azure CLI always starts with “az” as command (meaning Azure of course). The command for stopping one or more vm’s with the Azure CLI is “vm stop”. The only two things I need to provide now are the id’s I want to stop and “–nowait” since I want to quit the script right after.

So, the script would look like the following:

az vm stop --ids YOUR_IDS --no-wait

However, this has still one major disadvantage: you would need to provide all ID’s Hardcoded. This doesn’t matter at all if your cluster never changes, but in my case I add and delete vm’s to or from the cluster, so this script doesn’t play well for my case. However, the CLI is very flexible (and so is bash) and I can query all my vm’s in a resource group. This will give me the IDs which are currently in the cluster (let’s assume I delete dropped vm’s and add new vm’s to the RG). The Query for retrieving all VMs in a Resource Group is easy:

az vm list --resource-group YOUR_RESOURCE_GROUP --query "[].id" -o tsv

This will give me all IDs in the RG. The real fun starts when doing this in one statement:

az vm stop --ids \$(az vm list --resource-group clouderarg --query "[].id" -o tsv) --no-wait

Which is really nice and easy 🙂

It is similar with starting VMs in a Resource Group:

az vm start --ids \$(az vm list --resource-group mmhclouderarg --query "[].id" -o tsv) --no-wait

## International Data Science Conference, Salzburg

Hi,

I am happy to share this exciting conference I am keynoting at. Also, Mike Ohlsen from Cloudera will deliver a keynote at the conference.

June 12th – 13th 2017 | Salzburg, Austria | www.idsc.at

The 1st International Data Science Conference (iDSC 2017) organized by Salzburg University of Applied Sciences (Information Technology and Systems Management) in cooperation with Information Professionals GmbH seeks to establish a key Data Science event, providing a forum for an international exchange on Data Science technologies and applications.

The International Data Science Conference gives the participants the opportunity, over the course of two days, to delve into the most current research and up-to-date practice in Data Science and data-driven business. Besides the two parallel tracks, the Research Track and the Industry Track, on the second day a Symposium is taking place presenting the outcomes of a European Project on Text and Data Mining (TDM). These events are open to all participants.

Also we are proud to announce keynote presentations from Mike Olson (Chief Strategy Officer Cloudera), Ralf Klinkenberg (General Manager RapidMiner), Euro Beinat (Data-Science Professor and Managing Director CS Research), Mario Meir-Huber (Big Data Architect Microsoft). These keynotes will be distributed over both conference days, providing times for all participants to come together and share views on challenges and trends in Data Science.

The Research Track offers a series of short presentations from Data Science researchers on their own, current papers. On both conference days, we are planning a morning and an afternoon session presenting the results of innovative research into data mining, machine learning, data management and the entire spectrum of Data Science.

The Industry Track showcases real practitioners of data-driven business and how they use Data Science to help achieve organizational goals. Though not restricted to these topics only, the industry talks will concentrate on our broad focus areas of manufacturing, retail and social good. Users of data technologies can meet with peers and exchange ideas and solutions to the practical challenges of data-driven business.

Futhermore the Symposium is organized in collaboration with the FutureTDM Consortium. FutureTDM is a European project which over the last two years has been identifying the legal and technical barriers, as well as the skills stakeholders/practitioners lack, that inhibit the uptake of text and data mining for researchers and innovative businesses. The recommendations and guidelines recognized and proposed to counterbalance these barriers, so as to ensure broader TDM uptake and thus boost Europe’s research and innovation capacities, will be the focus of the Symposium.

Our sponsors ClouderaF&F and um etc. will have their own, special platform: half-day workshops to provide hands-on interaction with tools or to learn approaches to developing concrete solutions. In addition, there will be an exhibition of the sponsors’ products and services throughout the conference, with the opportunity for the participants to seek contact and advice.

The iDSC 2017 is therefore a unique meeting place for researchers, business managers, and data scientists to discover novel approaches and to share solutions to the challenges of a data-driven world.

There are several things people discuss when it comes to Hadoop and there are some wrong discussions. First, there is a small number of people believing that Hadoop is a hype that will end at some point in time. They often come from a strong DWH background and won’t accept (or simply ignore) the new normal. But there are also some people that basically coin two major sayings: the first group of people states that Hadoop is cheap because it is open source and the second group of people states that Hadoop is expensive because it is very complicated. (Info: by Hadoop, I also include Spark and alike)

Neither the one nor the other is true.

Now, we have the opposite. Hadoop is expensive. Is it? In the past years I saw a lot of Hadoop projects the went more or less bad. Costs were always higher than expected and the project timeframe was never kept. Hadoop experts have a high income as well, which makes consulting hours even more expensive. Plus: you probably won’t find them on the market, as they can select what projects to make. So you have two major problems: high implementation cost and low ressource availability.

Another factor that is relevant to the cost discussion is the cluster utilization. In many projects I could see one trend: when the discussion about cluster sizing is on, there are two main decisions: (a) sizing the cluster to the highest expected utilization or (b) making the cluster smaller than the highest expected utilization. If you select (a), you have another problem: the cluster might be under-utilized. What I could see and what my clients often have, is the following: 20% of the time, they have full utilization on the cluster, but 80% of the time the cluster utilization is below 20%. This basically means that your cluster is very expensive when it comes to business case calculation. If you select (b), you will loose business agility and your projects/analytics might require long compute times.

At the beginning of this article, I promised to explain that Hadoop is still cost-effective. So far, I only stated that it might be expensive, but this would mean that it isn’t cost effective. Hadoop is still cost effective but I will give you a solution in my next blog post on that, so stay tuned 😉

## New: Datascience with Apache Pig E-book for 0.99 cent instead of 5\$!

I am happy to announce that I’ve created a new e-book for Amazon Kindle. As a promotional offer, the e-book will only cost 0.99 cent the next 6 days and the price will then go up again to it’s original price tag! Make sure to obtain it now 🙂

You can obtain the e-book here.

## My Big Data predictions for 2016

As 2016 is around the corner, the question is what this year will bring for Big Data. Here are my top assumptions for the year to come:

• The growth for relational databases will slow down, as more companies will evaluate Hadoop as an alternative to classic rdbms
• The Hadoop stack will get more complicated, as more and more projects are added. It will almost take a team to understand what each of these projects does
• Spark will lead the market for handling data. It will change the entire ecosystem again.
• Cloud vendors will add more and more capability to their solutions to deal with the increasing demand for workloads in the cloud
• We will see a dramatic increase of successful use-cases with Hadoop, as the first projects come to a successful end

What do you think about my predictions? Do you agree or disagree?