Data science and machine learning are two hot topics right now and you might be interested in learning about them.
Many beginners will wonder whether they should start out by learning data science or machine learning and this post will try to help you with that.
So, should I learn machine learning or data science first? If your goal is to become a datascientist, it would be best to start by learning skills such as data cleaning, processing and analysis using things such as the Pandas library as a part of a data science course.
There are actually a lot of things to consider when deciding on whether to enter the machine learning or data science field. Additionally, even though the two fields are very related there are a number of key differences between them.
Machine learning is where computers learn from data and use that data to make predictions without being explicitly told how to. The machine learning algorithms are also able to adjust the predictions that they make when they are given new data and some of the algorithms are able to be used to find patterns in the data that humans wouldn’t normally be able to.
Examples of how machine learning can be used would include:
- Recommending videos on Youtube or Netflix based on a users watch history and the watch history of other users
- Recommending products on websites based on the purchase history of the person
- Predicting the price of a house based on data about things such as the number of floors or bedrooms and the prices of other houses in the area
Machine learning is also applicable in a wide range of different fields including:
You can watch the video below to see more about what machine learning is:
Machine learning jobs
There are a number of different machine learning based jobs that you can get and they include:
- Machine learning engineer
The role of a machine learning engineer is to develop and to deploy machine learning models at scale.
Typically, machine learning engineer jobs will require a masters degree in a field such as computer science or statistics. However, some people have been able to get jobs in the field with a bachelors degree by showing a lot of relevant experience. I have spoken about how you can get relevant experience, in the past, here.
According to Payscale, the median pay for a machine learning engineer is $110,000, the 10th percentile makes $76,000 and the 90th percentile makes $152,000.
You can watch an interview with a machine learning engineer below:
- Natural language processing researcher
A natural language processing researcher works on ways to improve products that involve language. Examples could include working on search autocomplete, home-assistants or translation.
Typically, NLP jobs will require that you have a Phd in a quantitative field. According to Payscale, the average salary for people with skills in NLP is $108,000.
- Computer vision researcher
A computer vision engineer will work on things that involve working with visual data. Examples of where computer vision jobs are used include self-driving cars, facial recognition and healthcare.
Computer vision jobs will often require that you have a Phd. According to Payscale, the average pay for a computer vision engineer is $91,000, the 10th percentile makes $74,000 and the 90th percentile makes $163,000.
How to learn machine learning and necessary skills
To learn machine learning it will be necessary for you to have a number of skills. These skills include knowledge of linear algebra, calculus, probability, statistics and programming.
There are actually some courses that will teach you machine learning that don’t assume any prior knowledge. One of those courses is the most popular machine learning course available right now which is taught by Andrew Ng from Stanford University. You can find the course here. I would recommend that you start with this course so that you can see whether or not machine learning is for you.
Once you have taken that course and you have decided that you are interested in persuing machine learning then it would be worthwhile for you to learn the required mathematics in order to fully understand the algorithms and how to statistically use them.
The courses that I would recommend include:
Linear algebra (The University of Texas at Austin)
Programming in Python (MIT)
Once you have learned the above then I would recommend Deep learning and machine learning (MIT). In this course, you will be able to learn the mathematical details of the machine learning and deep learning algorithms.
It would also be worthwhile for you to go through this book Hands on Machine Learning since it gives a very good overview of how to implement the machine learning algorithms in Python.
If you want to get a job in machine learning then it will also be necessary for you to learn about databases and computational complexity as well.
Data science is the process of organizing, analyzing and helping people to make decisions based on large amounts of data.
Machine learning is a key part of the data science process. However, most of the work that data scientists do goes into other areas of the data science process which is:
- Acquiring and storing data
- Asking about how that data might be useful
- Cleaning the data
- Doing exploratory data analysis which is where you summarize the main characteristics of the data
- Choosing and applying machine learning models to the data
- Making sense of the results of the models and how accurate they are
- Making decisions based on the results of the ML models
You can watch the video below to see more about what data science involves:
Data science jobs
Jobs in data science are currently high in demand and the demand for data science jobs is expected to rise, at a faster rate than the supply of workers, in the coming years (source).
According to Payscale, a data scientist will make $91,000 on average, the 10th percentile makes $62,000 and the 90th percentile makes $131,000.
Typically, it will be necessary to have a masters degree in a quantitative field to get a job in data science. However, there are some postings for people with just a bachelors degree and the ability to show that you have relevant experience. I have talked about how you can get that relevant experience, in the past, here.
How to learn data science and skills that are required
To learn data science it will be necessary for you to learn machine learning so I would recommend that you follow the same steps that I advised above to learn machine learning.
It will also be necessary for you to have a very good understanding of data analytics. I would recommend that you start out by watching this Youtube series which shows you how to do data analytics in Python.
If you don’t mind spending some money then I would recommend working through the material on the website Datacamp.com which will take you through the whole data science process.
How data science and machine learning are different
Even though a lot of what get done in machine learning and data science are similar, they are not the same thing.
The role of a data scientist will be to use data to help the business make better decisions and the use of machine learning will often help in doing this.
Whereas, the role of machine learning is to learn from data and to make predictions based on what it learns from the data.
Data science will usually be used in a business setting but work in machine learning can be used in a wide range of settings and there are many research opportunities in machine learning.
How machine learning engineers are different to data scientists
Two very similar job roles are that of a machine learning engineer and a data scientist.
The main difference between the two will normally be that machine learning engineers will focus on building and making machine learning models that are useable at scale. Whereas, a data scientist will generally be responsible for finding ways to use data to improve the workings of a business.
I have written more about how machine learning engineers and data scientists are different, in the past, here.
Why you should start out by learning data analytics
In both, machine learning and data science, it will be necessary for you to do a lot of data analytics and pre-processing.
This is so that you can make sense of what the data is showing, so that you can modify the data so that it works effectively with the machine learning models and so that you can remove unnecessary features in the data.
Additionally, when you start learning machine learning, many of the materials will assume that you have knowledge of how to do data analytics in a certain programming language (usually Python).