Hands-on Big Data & Deep Learning Training

Enabling learning through practical experimentation in the laboratory environment for the first time in the UK

Learn more
Image Image Image

Why choose DeepCI training courses?

DeepCI is offering unique range of unmatched hands-on training courses for developers, data scientists, and researchers, who share our passion to exploit the power of deep learning, optimization, and Big Data to solve a wide range of real-world problems.

DeepCI training covers both basic and advanced pre-processing (audio, video, text, EEG etc.), optimization, Big Data, and deep learning algorithms ranging from multilayer perceptrons, to convolutional neural network (CNN), long short-term memory (LSTM), and CNN-LSTM for both 2D/3D classification and regression problems, using popular deep learning tools including TensorFlow, DIGITS, Caffe, Keras, PyTorch, Theano, MATLAB.

Comparison with existing university/online available courses


University courses

Existing online courses


Hands-on training



State-of-the-art tools


Advanced big data and deep learning algorithms

Wide range of real-world applications

*Limited hands-on training
Out-dated tools
Limited range of applications, mostly focus on toy problems

Who are our courses for?

PhD researchers (prospective and current)
Industry practitioners struggling to formulate deep learning solutions to data-driven problems
Graduates looking for Data Science jobs
  • employees struggling to grasp deep learning concepts

Why deep learning?

Deep learning has been recognised as one of the ten breakthrough technologies according to MIT technology review. Today deep learning is one of the hottest and fastest-growing areas of research in AI and machine learning. World leading high-tech companies including Microsoft, Google, Facebook, IBM, Apple and many more, are continuing to invest billions of dollars in deep learning research, for pioneering next-generation data-driven solutions to a whole range of challenging applications. The latter include automatic pattern recognition, image segmentation, classification, object detection and tracking, intention modelling, deception detection, data and cyber-security, healthcare, remote sensing, quantum and cloud computing, and lots more!

Deep learning uses multi-layered deep neural networks (DNNs) to automatically learn multiple levels of abstraction and representation that can extract and reveal hidden information, patterns, trends and correlations in heterogeneous Big Data, such as very high resolution images, sound, and text.

Hands-on training and understanding of cutting-edge deep learning technologies can open up untapped market opportunities for both existing businesses and entrepreneurs/startups of tomorrow!

Beginners/Intermediate-level Hands-on Training

Pre-requisites: None

Participants will learn:

Foundations of deep learning and state-of-the-art deep neural network architectures.

Innovative tricks to optimise the performance of Deep Learning architectures and solutions

Build, train, test, and fine-tune fully-connected deep neural networks for real-world case studies (participants will be able to select from a wide-range of benchmark data-driven applications).

Course content:

  • Classification vs Regression
  • Linear regression
  • Logistic regression
  • Vectorization
  • Gradient descent
  • Softmax regression
  • Bias and variance
  • Multi-layer artificial neural network
  • Supervised neural network
  • Why deep neural network?
  • Importance of pre-processing
  • Big Data pre-processing (Audio, Visual, Text, EEG etc.)
  • Feature extraction using convolution
  • Pooling
  • Dropout regularization
  • L1/L2 regularization
  • Weight initialization for deep networks
  • Mini-batch gradient descent
  • Convolutional neural networks for classification/recognition problems
  • Image, audio, video, and other 2D/3D data.
  • IMDB movie reviews sentiment, MNIST, CIFAR 10, CIFAR 100, CT scan, MRI images etc.
  • Convolutional neural network for regression/sequential/time-series problems
  • Image, audio, video, and other 2D/3D data.
  • Appliances energy prediction, air quality, wireless QoS prediction etc.
  • Recurrent neural networks for classification/recognition problems
  • Image, audio, video, and other 2D/3D data.
  • MNIST, CIFAR 10, CIFAR 100, CT scan, MRI images
  • Image, audio, video, and other 2D/3D data.
  • Deep autoencoders for compression, dimensionality reduction, and denoising 
  • The next Beginners/Intermediate course is scheduled on 16-17 June 2018, limited places - register your interest and reserve your place now

    Advanced and application specific training course

    Pre-requisites: These courses are specifically designed for participants with good neural network/Data Science background.

    Participants will learn:

    How to comprehend, define, and formulate any real-world problem

    How to define efficient context-aware strategy taking into account the
    application requirements and constraints comprising trade-offs among computational complexity, processing time, available dataset etc.

    Smart integration of machine learning and optimization algorithms such as artificial neural network and genetic algorithm

    Selection of optimal machine learning, training, and reasoning algorithms
    Concepts such as multimodal, context-aware, brain-inspired

    Hands-on problem solving applications (not limited to)

  • Fraud detection
  • Brand reputation management
  • Fake news prediction
  • Self-driving cars
  • Equipment maintenance and diagnostic tools
  • 5G wireless solutions
  • Climate change prediction
  • Smart meters
  • Smart grids
  • Deception detection
  • Emotion recognition
  • Cancer diagnosis (Breast/Lungs/Liver
  • Speech recognition
  • Speech enhancement
  • Sentiment analysis 

  • The next Advanced and application specific course is scheduled on 23-24 June 2018, limited places - register your interest and reserve your place now