Fisher kernel for tree structured data

Nicotra, Luca; Micheli, Alessio; Starita, and Antonina

doi:10.1109/ijcnn.2004.1380905

Cited by 16 publications

(5 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Deriving a probabilistic model incorporating information about data, however, is in general computationally very expensive. These models include [58][59][60][61][62][63]. A very important kernel in this class is the Fisher kernel [61], which actually is a general way on how to generate a kernel using a probabilistic model describing the data, such as a Hidden Markov Model (used, for example, in Bioinformatics to model protein families) or Stochastic Context-Free Grammars (which have been used to model RNA sequences).…”

Section: Kernel Methodsmentioning

confidence: 99%

An Introduction to Recursive Neural Networks and Kernel Methods for Cheminformatics

Micheli¹,

Sperduti²,

Starita

2007

CPD

View full text Add to dashboard Cite

show abstract

Section: Kernel Methodsmentioning

confidence: 99%

An Introduction to Recursive Neural Networks and Kernel Methods for Cheminformatics

Micheli¹,

Sperduti²,

Starita

2007

CPD

View full text Add to dashboard Cite

show abstract

“…In this thesis the Fisher kernel [15] is explained since it has been shown to perform well in many applications and can process data that are not of the vector type [16]. Generative models such as Gaussian Mixture Models (GMM) and Hidden Markov Models (HMM) [17] have been vastly used to model the observed data.…”

Section: Fisher Kernelmentioning

confidence: 99%

Predicting system collapse : application of kernel-based machine learning and inclination analysis

Hosseinizadeh¹

2021

Preprint

View full text Add to dashboard Cite

While many modelling methods have been developed and introduced to predict the actual state of a system at the next point of time, the purpose of this research is to present and discuss two approaches NOT to predict the exact future states, but to identify the potential for final collapse of a system. The first approach is based on kernel methods, a sub category of supervised learning, and attempts to provide a visualization method to classify the active and dead companies and predict the potential collapse of a system. The second method aims to analyze the inclination of a system by looking at the local changes that have been observed over a certain period of time in the past. Application of these modelling approaches to predict collapse in different companies belonging to two industrial sectors by looking at behaviour of their closing stock prices are discussed in this research. Advantages and limitations of each approach are also discussed.

show abstract

“…-text (via NLP parse-tree kernels/LSA kernels [9,10]) -graphs and sequences (via string kernels, random walk kernels [11,12]) -shapes (via edit distance kernels [13]) -real vectors (via dot product kernels, polynomial kernels etc [14]) -sets of pixels/voxels (efficient match kernels, pyramid match kernels [15]) -stochastic data (Fisher kernels [8])…”

Section: Mercer Kernels Are Now For All Data Typesmentioning

confidence: 99%

A Kernel-Based Framework for Medical Big-Data Analytics

Windridge

Bober

2014

Interactive Knowledge Discovery and Data Mining in Biomedical Informatics

View full text Add to dashboard Cite

Abstract. The recent trend towards standardization of Electronic Health Records (EHRs) represents a significant opportunity and challenge for medical big-data analytics. The challenge typically arises from the nature of the data which may be heterogeneous, sparse, very highdimensional, incomplete and inaccurate. Of these, standard pattern recognition methods can typically address issues of high-dimensionality, sparsity and inaccuracy. The remaining issues of incompleteness and heterogeneity however are problematic; data can be as diverse as handwritten notes, blood-pressure readings and MR scans, and typically very little of this data will be co-present for each patient at any given time interval.We therefore advocate a kernel-based framework as being most appropriate for handling these issues, using the neutral point substitution method to accommodate missing inter-modal data. For pre-processing of image-based MR data we advocate a Deep Learning solution for contextual areal segmentation, with edit-distance based kernel measurement then used to characterize relevant morphology.

show abstract

Fisher kernel for tree structured data

Cited by 16 publications

References 17 publications

An Introduction to Recursive Neural Networks and Kernel Methods for Cheminformatics

An Introduction to Recursive Neural Networks and Kernel Methods for Cheminformatics

Predicting system collapse : application of kernel-based machine learning and inclination analysis

A Kernel-Based Framework for Medical Big-Data Analytics

Contact Info

Product

Resources

About