About this Course
Within the second course of Machine Studying Engineering for Manufacturing Specialization, you’ll construct knowledge pipelines by gathering, cleansing, and validating datasets and assessing knowledge high quality; implement characteristic engineering, transformation, and choice with TensorFlow Prolonged and get probably the most predictive energy out of your knowledge; and set up the information lifecycle by leveraging knowledge lineage and provenance metadata instruments and comply with knowledge evolution with enterprise knowledge schemas.
WHAT YOU WILL LEARN
Establish accountable knowledge assortment for constructing a good ML manufacturing system.
Implement characteristic engineering, transformation, and choice with TensorFlow Prolonged
Perceive the information journey over a manufacturing system’s lifecycle and leverage ML metadata and enterprise schemas to deal with shortly evolving knowledge.
SKILLS YOU WILL GAIN
- ML Metadata
- Convolutional Neural Community
- TensorFlow Prolonged (TFX)
- Information Validation
- Information transformation
Syllabus – What you’ll study from this course
7 hours to finish
Week 1: Accumulating, Labeling and Validating Information
This week covers a fast introduction to machine studying manufacturing programs. Extra concretely you’ll find out about leveraging the TensorFlow Prolonged (TFX) library to gather, label and validate knowledge to make it manufacturing prepared.
7 hours to finish
Week 2: Characteristic Engineering, Transformation and Choice
Implement characteristic engineering, transformation, and choice with TensorFlow Prolonged by encoding structured and unstructured knowledge varieties and addressing class imbalances
5 hours to finish
Week 3: Information Journey and Information Storage
Perceive the information journey over a manufacturing system’s lifecycle and leverage ML metadata and enterprise schemas to deal with shortly evolving knowledge.
3 hours to finish
Week 4 (Elective): Superior Labeling, Augmentation and Information Preprocessing
Mix labeled and unlabeled knowledge to enhance ML mannequin accuracy and increase knowledge to diversify your coaching set.
0 Comments