Este curso forma parte de Programa especializado: Informed Clinical Decision Making using Deep Learning

Fani Deligianni

Ofrecido Por

Programa especializado: Informed Clinical Decision Making using Deep LearningUniversity of Glasgow

Acerca de este Curso

2262 vistas recientes

This course will introduce the concepts of interpretability and explainability in machine learning applications. The learner will understand the difference between global, local, model-agnostic and model-specific explanations. State-of-the-art explainability methods such as Permutation Feature Importance (PFI), Local Interpretable Model-agnostic Explanations (LIME) and SHapley Additive exPlanation (SHAP) are explained and applied in time-series classification. Subsequently, model-specific explanations such as Class-Activation Mapping (CAM) and Gradient-Weighted CAM are explained and implemented. The learners will understand axiomatic attributions and why they are important. Finally, attention mechanisms are going to be incorporated after Recurrent Layers and the attention weights will be visualised to produce local explanations of the model.

Fechas límite flexibles

Restablece las fechas límite en función de tus horarios.

Certificado para compartir

Obtén un certificado al finalizar

100 % en línea

Comienza de inmediato y aprende a tu propio ritmo.

Curso 3 de 5 en

Programa especializado: Informed Clinical Decision Making using Deep Learning

Nivel intermedio

Python programming and experience with basic packages such as numpy, scipy and matplotlib

Aprox. 30 horas para completar

Inglés (English)

Qué aprenderás

Program global explainability methods in time-series classification
Program local explainability methods for deep learning such as CAM and GRAD-CAM
Understand axiomatic attributions for deep learning networks
Incorporate attention in Recurrent Neural Networks and visualise the attention weights

Habilidades que obtendrás

attention mechanisms
explainable machine learning models
model-agnostic and model specific models
global and local explanations
interpretability vs explainability

Fechas límite flexibles

Restablece las fechas límite en función de tus horarios.

Certificado para compartir

Obtén un certificado al finalizar

100 % en línea

Comienza de inmediato y aprende a tu propio ritmo.

Curso 3 de 5 en

Programa especializado: Informed Clinical Decision Making using Deep Learning

Nivel intermedio

Python programming and experience with basic packages such as numpy, scipy and matplotlib

Aprox. 30 horas para completar

Inglés (English)

Instructor

Fani Deligianni

School of Computing Science

1877 alumnos

5 cursos

Ofrecido por

University of Glasgow

The University of Glasgow has been changing the world since 1451. It is a world top 100 university (THE, QS) with one of the largest research bases in the UK.

We are a member of the prestigious Russell Group of leading UK Universities with annual research income of more than £179m. The University’s #TeamUofG community is truly international with over 8000 staff and 28,0000 students from more than 140 countries. A 2019 Time Out survey placed Glasgow in the top ten cities in the world. Ranked between Berlin and Paris, Glasgow was voted number one for both friendliness and affordability. Right now our dedicated community of staff, students and alumni is working to address the challenges of Covid-19 and understand how we can make life safer for everyone.

Programa - Qué aprenderás en este curso

Semana1

Semana 1

9 horas para completar

Interpretable vs Explainable Machine Learning Models in Healthcare

Deep learning models are complex and it is difficult to understand their decisions. Explainability methods aim to shed light to the deep learning decisions and enhance trust, avoid mistakes and ensure ethical use of AI. Explanations can be categorised as global, local, model-agnostic and model-specific. Permutation feature importance is a global, model agnostic explainabillity method that provide information with relation to which input variables are more related to the output.

9 horas para completar

6 videos (Total 72 minutos), 8 lecturas, 1 cuestionario

6 videos

Welcome video - Explainable Deep Learning Models for Healthcare55sInterpretability vs Explainability16m'Explainability' in Healthcare Applications17mTaxonomy of Explainability Methods17mModel Agnostic Explainability Methods10mPermutation Feature Importance in Time Series Data8m

8 lecturas

The importance of explainable prediction models in healthcare10mExplainable Artificial Intelligence - Taxonomy1hModel Agnostic Explainability30mPermutation Feature Importance1hPractical Exercise: Interpretability of the MLP model using Permutation Feature Importanceg1hPractical Exercise: Interpretability of the CNN model using Permutation Feature Importance1hPractical Exercise: Interpretability of the LSTM model using Permutation Feature Importance1hExplainability models in ECG50m

1 ejercicio de práctica

End of week 1 quiz30m

Semana2

Semana 2

8 horas para completar

Local Explainability Methods for Deep Learning Models

Local explainability methods provide explanations on how the model reach a specific decision. LIME approximates the model locally with a simpler, interpretable model. SHAP expands on this and it is also designed to address multi-collinearity of the input features. Both LIME and SHAP are local, model-agnostic explanations. On the other hand, CAM is a class-discriminative visualisation techniques, specifically designed to provide local explanations in deep neural networks.

8 horas para completar

5 videos (Total 48 minutos), 7 lecturas, 1 cuestionario

5 videos

Local Interpretable Model Agnostic Explanations (LIME)9mLIME in Time-Series Classification9mShapley Additive Explanations11mModel-Specific Explanations: Visualisation Methods12mCAM in Time-Series Classification4m

7 lecturas

Why Should I Trust You?30mPractical Exercise: Interpretability of heartbeat classification using LIME and an NNMLP model1hPractical Exercise: Interpretability of heartbeat classification using LIME and a CNN model1hPractical Exercise: Interpretability of heartbeat classification using LIME and an LSTM model1hA Unified Approach to Interpreting Model Predictions40mPractical Exercise: Interpretability of CNN models using Class Activation Maps10mClass Activation Mapping1h

1 ejercicio de práctica

End of week 2 quiz30m

Semana3

Semana 3

8 horas para completar

Gradient-weighted Class Activation Mapping and Integrated Gradients

8 horas para completar

4 videos (Total 37 minutos), 6 lecturas, 1 cuestionario

4 videos

Gradient Weighted Class Activation Maps11mGrad-CAM in Time-Series Classification8mIntegrated Gradients12mIntegrated Gradients in Time Series Classification3m

6 lecturas

GRAD - Class Activation Mapping 30mPractical Exercise: Interpretability of the CNN model using Gradient-weighted Class Activation Mapping1hPractical Exercise: Interpretability of the LSTM model using Gradient-weighted Class Activation Mapping1hAxiomatic Attribution for Deep Networks40mPractical Exercise: Interpretability of the CNN model using Integrated Gradients1hPractical Exercise: Interpretability of the LSTM model using Integrated Gradients1h

1 ejercicio de práctica

End of week 3 quiz30m

Semana4

Semana 4

5 horas para completar

Attention mechanisms in Deep Learning

Attention in deep neural networks mimics human attention that allocates computational resources to a small range of sensory input in order to process specific information with limited processing power. In this week, we discuss how to incorporate attention in Recurrent Neural Networks and autoencoders. Furthermore, we visualise attention weights in order to provide a form of inherent explanation for the decision making process.

5 horas para completar

3 videos (Total 34 minutos), 3 lecturas, 2 cuestionarios

Acerca de Programa especializado: Informed Clinical Decision Making using Deep Learning

This specialisation is for learners with experience in programming that are interested in expanding their skills in applying deep learning in Electronic Health Records and with a focus on how to translate their models into Clinical Decision Support Systems.

The main areas that would explore are: Data mining of Clinical Databases: Ethics, MIMIC III database, International Classification of Disease System and definition of common clinical outcomes. Deep learning in Electronic Health Records: From descriptive analytics to predictive analytics Explainable deep learning models for healthcare applications: What it is and why it is needed Clinical Decision Support Systems: Generalisation, bias, ‘fairness’, clinical usefulness and privacy of artificial intelligence algorithms.

Informed Clinical Decision Making using Deep Learning

Preguntas Frecuentes

¿Cuándo podré acceder a las lecciones y tareas?
El acceso a las clases y las asignaciones depende del tipo de inscripción que tengas. Si tomas un curso en modo de oyente, verás la mayoría de los materiales del curso en forma gratuita. Para acceder a asignaciones calificadas y obtener un certificado, deberás comprar la experiencia de Certificado, ya sea durante o después de participar como oyente. Si no ves la opción de oyente:
es posible que el curso no ofrezca la opción de participar como oyente. En cambio, puedes intentar con una Prueba gratis o postularte para recibir ayuda económica.
Es posible que el curso ofrezca la opción 'Curso completo, sin certificado'. Esta opción te permite ver todos los materiales del curso, enviar las evaluaciones requeridas y obtener una calificación final. También significa que no podrás comprar una experiencia de Certificado.
¿Qué recibiré si me suscribo a este Programa especializado?
Cuando te inscribes en un curso, obtienes acceso a todos los cursos que forman parte del Programa especializado y te darán un Certificado cuando completes el trabajo. Se añadirá tu Certificado electrónico a la página Logros. Desde allí, puedes imprimir tu Certificado o añadirlo a tu perfil de LinkedIn. Si solo quieres leer y visualizar el contenido del curso, puedes auditar el curso sin costo.
¿Hay ayuda económica disponible?
Sí. En ciertos programas de aprendizaje, puedes postularte para recibir ayuda económica o una beca en caso de no poder costear los gastos de la tarifa de inscripción. Si hay ayuda económica o becas disponibles para tu selección de programa de aprendizaje, verás un enlace para postularte en la página de descripción.

¿Tienes más preguntas? Visita el Centro de Ayuda al Estudiante.

Explainable deep learning models for healthcare - CDSS 3

Acerca de este Curso

Qué aprenderás

Habilidades que obtendrás