Machine Learning Course

siemens intel tcs cognizant virtusa accenture nference hcl

Customized Machine Learning training

Does your team regularly work with machine learning algorithms and models? If so, we have just the thing for you! Tailored training courses are also available for machine learning specialists and developers. Optimize your AI projects to the highest level. Thanks to individual modules, your data scientist and AI teams can receive customized training.

We are sure to find the optimal combination for your team's needs. Bring all your questions. We are happy to take the time to advise you individually.

All courses live

In the Academy or online

Your Team Benefits at a Glance

  • Customized Topics

  • Flexible Dates and Training Formats

  • Experienced Trainers

  • Certificate for Active Participation

  • Learning Material

  • Practice-oriented, Compact and Lively

Machine learning topics and modules

Just to clarify upfront: Any topic can be explored in depth at your request. Your needs come first. You decide what your team needs. We create the concept. If your team just wants an overview, then you can select a wide range of topics. However, it will likely be challenging for the team to implement what they have learned in practice afterwards.

So, if you want to work seriously and professionally with Python, it's better, based on our experience, to select fewer topics which we then discuss in depth during the training and conduct a sufficient number of practical workshops. Here, indeed, "less is more". In a consultation, we can best jointly identify which topics will be most beneficial for your team and your projects.

We don't preach theory, we solve problems! Get the maximum knowledge for your team. Compile the topics according to your previous knowledge and needs, depending on the project. Because individual learning brings the desired success!

Machine learning techniques

In very simplified terms, classical techniques are well suited for tabular or structured data, while deep learning techniques are better suited for problems that process image, text, or speech data.

With Deep Learning techniques it is also state of the art not to develop the models from scratch but to use so-called pre-trained models. For most use cases we recommend to focus the course on the use of pre-trained models and not on the creation of own models.

Basics and traditional techniques

If your team has no prior experience with machine learning techniques, it is recommended that you discuss some of the basics and traditional techniques using relatively simple examples so that you have a solid foundation for working on more complex problems later.

  • Types of Machine Learning
  • The machine learning process (for supervised learning).
  • Linear and logistic regression
  • Preparation of data
  • Visualization of data
  • Decision trees, random forests, gradient boosted trees
  • Designing a training process
  • Finding errors (what to do if the training does not converge)
  • Unsupervised Learning

Deep Learning Techniques

With the following topics, your team can take their knowledge of building deep learning models to the next level:

  • Fundamentals of Deep Learning and Neural Networks.
  • The Multilayer Perceptron (Fully Connected Net)
  • Convolutional Networks
  • Recurrent Networks (LSTM, GRU).
  • Improvement of network architectures
  • (skip connections, dropout, regularization,...)
  • Attention mechanisms
  • The Transformer Architecture

Pretrained models

For many practical problems, working with pre-trained models lends itself to:

  • Pretrained models for image processing.
  • Pre-trained models for natural language processing (NLP).
  • Hugging Face Hub, Transformers, and Datasets.
  • The Haystack Framework for Search Engines

Deep Learning Packages

The following list contains only a selection of the numerous Python libraries that you can use to solve machine learning problems. Typically, the selection of libraries discussed in a training session is based on the desired application areas and techniques, but it is of course also possible to conduct a training session on a specific Python library that your team wants to delve deeper into.

  • Basic libraries:
  • Mathematics/linear algebra: NumPy
  • Tabular Data: Pandas
  • Visualization: Matplotlib, Seaborn
  • SciPy (for statistics and scientific computing)
  • Statsmodels (for statistics and forecasting)
  • Scikit-Learn (for data preparation and machine learning)
  • XGBoost: (for machine learning with gradient boosted trees)
  • LightGBM: (similar to XGBoost, but more efficient for some applications)
  • Deep Learning/GPU Computing
  • TensorFlow
  • PyTorch
  • Transformers
  • Datasets
  • Haystack
  • Pmdarima
  • Streamlit for Dashboards

Working effectively as a team

These modules cover techniques that help make teamwork more productive and improve the quality of the resulting software.

The Working effectively as a team module includes the following focus areas:

  • Version control (Git)
  • Docker
  • Continuous Integration (CI)
  • Workflows: Issue Tracking, Pull/Merge Requests
  • Example: working with a local GitLab instance in Docker
  • Kubernetes
  • Elasticsearch/Kibana (OpenSearch/Dashboards)

Certificate

Of course, as a participant in a Machine Learning course, you will receive a certificate. The prerequisite for this is full participation in all course units and programming tasks, and the successful programming of a small final project. However, after an intensive Machine Learning course, this will certainly bring you more joy than stress.

Your Trainers

Dr. Arvind Khanna
Dr. Arvind Khanna

Expert: Machine Learning, Deep Learning, NLP, Python, TensorFlow, PyTorch, Data Visualization

Rajiv Sharma
Rajiv Sharma

Expert:Model Deployment, Kubernetes for ML, CI/CD for ML, Feature Stores, Model Monitoring & Optimization

Ankit Verma
Ankit Verma

Expert: Time Series Analysis, Regression Modeling, Bayesian Statistics, A/B Testing, R Programming, Hypothesis Testing

NO QUESTION REMAINS UNANSWERED HERE!

We are very happy to take time for you. Use our contact form for a written contact. Our team will respond quickly and within 24 hours at the latest. In case of technical questions and to clarify the focal points, which are target-oriented for your team, please use the possibility of a non-binding consultation appointment.