Index
- (Breeze's / Halflife) Utility
- accuracy
- Accuracy of predictions
- activation function
- actual
- adapt
- adaptively
- Adaptivity and learning
- Adjusted Cosine
- adjustment
- Alan Turing
- all-or-nothing theory
- All Movie Guide
- Amazon
- Amazon homepage
- approach
- approximation
- architectural graph
- Artificial neural network
- Artificial neural networks
- artist artist algorithm
- artist to artist
- assumes
- attribute models
- Average First Hit Position
- Average quantization error
- Average rating prediction
- Average User Coverage
- Average User F1
- Average User MAE
- Average User MSE
- Average User NMAE
- Average User Precision
- Average User Recall
- Average User RMSE
- axons
- backpropagation
- Backward propagation
- bag of words
- bags of words
- balance point
- BASELINESF Technique
- Batch Learning
- Bayesian clustering
- Bayesian Networks
- best matching node
- Better together
- Blood
- Boltzmann learning
- border effect
- Brain maps
- Browsing
- buddies
- case amplification
- CF using SOM and ART2 networks
- Challenges
- CIPUCF Technique
- closed form
- cognitive
- collaboration via content
- Collaborative Filtering
- collaborative tagging
- columns
- Communications of the ACM
- compatible
- competitive learning
- complex
- Conclusions
- conditional probability
- connection links
- Content-boosted CF
- Content Based Filtering
- content filters
- context
- continuous
- converged
- convergence phase
- Correctness
- correlation based
- Cosine angle
- cosine based
- cosine normalization
- Cosine similarity algorithm
- cost function
- Coverage
- create-map
- credit assignment problem
- current iteration number
- curvlinear component analysis
- customer
- Customer reviews and ratings
- Customer tagged this item with
- death
- default voting
- dendrites
- Density matching
- descriptive
- desired
- directed links
- discriminant
- Discshop rating dataset (DS)
- discussion forums
- Disembodied Hand
- dissimilarity
- distance matrix
- distances
- Does it work?
- economic
- edit movie information
- Electronic junk
- Elman network
- ephemeral recommendations
- error-correction learning
- error-surface
- error signal
- Euclidian distance
- Evaluation data
- Evaluation protocols
- Evaluation results
- Evil Dead II
- examining new data
- excitation
- explained
- Explanation of recommendations
- explanations
- explicitly
- Exploding head
- exploratory data analysis
- exponential growth
- F1
- FAB
- features
- feedback
- feedback connections
- feedforward
- feedforward connections
- Feedforward networks
- Feedforward pass
- film finder
- First rater problem
- fixed
- forgetting factor
- Frequent item recommendations
- frequent item sets
- Future work
- Fuzziness
- Gaussian function
- generalization
- generalize
- generalized delta
- global
- globally ordered maps
- global ordering
- global shape
- Goodness recommendation
- Gore
- gradient descent method
- Grey sheep problem
- GroupLens
- growing
- guassian neighborhood function
- halflife
- hard-limiter
- heavyside function
- Hebbian learning
- hexagonal
- hidden layers
- hierarchical
- hopefully
- Hopfield network
- How it works
- how they work
- How to do it?
- how well they work
- Hybrid approach
- Hybrid Filters
- ICF Technique
- identity
- if such a discriminant exists
- image
- IMDb attribute dataset (IMDB)
- implicitly
- improve
- Incomplete data
- Incremental Learning
- information
- information analysis
- information filtering
- information overload
- information retrieval
- inhibition
- initialized
- Initialize network
- input layer
- instantaneous value
- Interaction design
- Internet Movie Database
- inverse document frequency
- inverse of time
- inverse user frequency
- irregular
- ISOMUCF Technique
- Item attribute models
- item attributes
- item based CF
- Item mean
- Item mean prediction algorithm
- item profiles
- Item rating models
- items
- item set to item CF
- Item transaction models
- Jordan network
- justification
- K-means
- K-means algorithm
- lateral-feedback connections
- lateral feedback
- lattice
- learn
- learning
- learning-rate factor
- learning-rate parameter
- learning ability
- learning algorithm
- learning paradigm
- Learning paradigms
- learning process (algorithm)
- Learning protocols
- Learning Rules
- least mean square
- Limited Content Analysis
- linear
- linear combiner
- linearly separable
- Linear projection
- Listmania!
- local
- local shape
- logistic function
- long-termed user-profiles
- majority
- Manual approach
- Map navigation module
- Map stretching module
- Markov decision chains
- Massive parallelism
- McCulloch-Pitts model
- mean-square-error
- Mean Absolute Error (MAE)
- mean squared difference
- Mean Squared Error (MSE)
- Mexican hat
- model
- model vector
- modules
- momentum term
- most likely
- movie
- MovieLens
- MOVSOM - State of the art
- MOVSOM architecture
- MOVSOM interface
- MOVSOM map
- MPSOMUCF Technique
- multidimensional scaling
- multilayer perceptrons
- Music artist reviews SOM
- negatives
- neighborhood
- neighborhood based
- neighborhood based CF
- neighborhood function
- neighborhood set
- Neighbor prediction
- neighbors
- neurons
- neurotransmitters
- new
- new item
- New item problem
- New user problem
- node-distance
- nodes
- Non-linear projection
- nonlinear
- normalization
- normalized document lengths
- normalized term frequencies
- normalized term frequency
- normalize the input
- Note
- objective function
- old
- ordered
- ordering
- ordering phase
- Other features
- output layer
- P-TANGO
- parallel
- parallel computational systems
- Patents
- pattern recognition
- PCBF Technique
- Pearson similarity algorithm
- people finder
- perceptron
- perceptron convergence theorem
- persistent recommendations
- personal
- personalized
- Plasticity
- positive
- postsynaptic
- Precision
- predefined
- predicted ratings
- prediction
- Prediction algorithms
- Prediction evaluation metrics
- predictions
- Present input
- presynaptic
- principal component analysis
- principal curves
- probabilistic
- processing elements
- processing methods
- profile analysis
- Profile representation
- promotes
- Properties of the SOM
- purchased
- Quality measure of the SOM
- quality of predictions
- quantization error
- Quick Browse
- r
- random-vector
- Random prediction algorithm
- Ranking algorithms
- Ranking evaluation metrics
- ranking lists
- rating models
- Rating module
- Recall
- Recommendation approaches
- recommendation center
- Recommendation module
- Recommendation profiles
- Recommendations
- Recommendation techniques
- recommender system
- Recommender Systems
- rectangular
- recurrent
- Recurrent networks
- reference
- reference vector
- regardless
- reinforcement learning
- relaxation
- relevance
- relevance feedback
- reliability
- Repeat
- response time
- result
- Retraining the SOM
- RI
- Ringo
- RNICF Technique
- RNUCF Technique
- robotics
- Robustness and fault tolerance
- Root Mean Squared Error (RMSE)
- rows
- Rule based using Ripper
- Sammon's mapping
- Sammons's mapping
- sample-initialization
- satisfaction
- Scalability
- Searching
- Search module
- Selection of parameters
- Selection of the BMN
- self-feedback
- self-organizing (feature) map
- Self-Organizing Maps
- sequential decision process
- Serendipity problem
- shape
- short-termed user profiles
- sigmoid
- signal-flow graphs
- significance weighting
- similar item recommendations
- similarity
- Similarity matching
- Similarity measures
- Similar movies
- single
- singular value decomposition
- Site features
- size
- slightly
- smoothing
- social
- soma
- SOM based applications
- SOMGOODNESSCF Technique
- SOMICF Technique
- SOM implementation
- SOM module
- SOM Toolbox for Matlab
- SOMUCF Technique
- source nodes
- So you'd like to...
- Sparsity
- speech analysis
- Splitting evaluation data
- stability-plasticity dilemma
- Statistical approach
- step-function
- strength
- structure
- Summary and conclusions
- Summary of the SOM algorithm
- summation function
- supervised
- Supervised learning
- synapses
- synaptic weights
- Syskill Webert
- tag
- Tagging module
- TBTOPNICF Technique
- TBTOPNUCF Technique
- techniques
- term-weighting
- term discrimination
- term frequency
- test data
- Text REtrieval Conference
- tf-idf
- The artificial neuron
- The batch SOM algorithm
- The biological neuron
- The brain
- The Dot-Product SOM algorithm
- The incremental SOM algorithm
- The Learning rate parameter
- The learning steps
- The Mexican Hat
- The Neighborhood function
- The page You made
- The Size of SOM
- The size of the Neighborhood
- The SOM paradigm
- The Thing
- threshold value
- time
- top N
- TOPNCBF Technique
- TOPNICF Technique
- top n item recommendations
- TOPNPCBF Technique
- TOPNSOMGOODNESSCF Technique
- TOPNSOMICF Technique
- TOPNSOMUCF Technique
- TOPNUCF Technique
- Topographic error
- topographic map
- Topological ordering
- totally
- training data
- transaction models
- transitivity
- transparency
- trust
- Trust in recommender systems
- TRUSTUCF Technique
- trustworthiness
- UCF Technique
- union
- unsupervised
- Unsupervised learning
- Update
- user based
- user based CF
- User mean prediction algorithm
- user profiles
- User rating models
- user to user
- User transaction models
- vector projection
- vector quantization
- vector quantization-projection
- vector space model
- video@bellacore
- Visualizing the SOM
- VISUALLY
- Visual recommendations
- Voroni tessellation
- weight
- Weighted average rating
- weighted deviation from mean
- Weighted sum
- Where are we today?
- Windrow-Hoffs delta rule
- winner-takes-all neurons
- Your community
- Your recent history
- Your Store
- Zombie