New machine learning specialization on Coursera teaches you to build production-ready models on GCP
By Rochana Golani, Director, Google Cloud Learning and Enablement
According to LinkedIn’s 2017 U.S. Emerging Jobs Report, demand for machine learning engineers has grown nearly tenfold since 2012, making it the top emerging job type. If you've been wondering how to best prepare for the opportunities that machine learning is just beginning to surface, you're not alone.
After the release of Google's Machine Learning Crash Course, we quickly realized that developers, data engineers, and data scientists were more than ready for deeper, contextualized learning and skills they could apply on the job. To address their needs, we have developed a series of deep-dive courses to help enthusiastic learners—who are oftentimes working professionals—build production-ready machine learning models that scale.
We are launching this set of five courses, based on the same content that has trained thousands of Google engineers, as a machine learning specialization on Coursera—Machine Learning with TensorFlow on Google Cloud Platform marks our seventh specialization, reinforcing our commitment to make high quality instruction widely accessible to everyone. We are also strong believers that in order to realize enduring, technical, and cloud-first skills, experiential learning is a must. That’s why we focused on injecting real-world, hands-on examples throughout these courses. By the end, you will be able to train practical machine learning models on real world data and deploy them in production. You'll be able to apply them to current data science challenges you encounter at work, to expand your current on-the-job technical skills, or to present new skills to an interviewer.
In this new on-demand series of courses, you’ll practice every aspect of machine learning, from problem formulation and data exploration, to training and performance tuning. You’ll also learn how to evaluate and deploy machine learning models from Google experts who share tricks and techniques to creatively overcome common problems that arise when training or deploying machine learning models in production. At the end of this specialization, you will have acquired a repertoire of skills along with a better understanding of the capabilities of TensorFlow, Cloud ML Engine, Cloud Datalab, Cloud Dataflow, Google BigQuery, Pandas, and NumPy. Through faster training speed, interactive tuning, and efficient inference, you’ll experience first-hand how Google Cloud Platform has been designed for real-world, scalable machine learning.
In the 5-course journey, you can:
- Learn the five phases of developing a candidate use case into a solvable machine learning problem
- Create ML-oriented datasets that are conducive to successful training runs, as well as higher accuracy and recall metrics, particularly under real-world classification and estimation conditions
- Write distributed machine learning models in TensorFlow: for example, you’ll learn to scale the model training process to utilize multiple workers
- Improve the accuracy of ML models through feature engineering
- Tune machine learning models to find the right mix of (hyper-)parameters that yields accurate, generalized models
Where to begin:
On-demand courses on Coursera make learning ML easy and accessible, and include presentations, self-paced labs, and videos. To find out more about the courses, join us for a 45-minute webinar where we’ll introduce this specialization and also provide open access to the first course, How Google does Machine Learning.