Neuroscience background
deep learning-recurrent structure
Neuroscience background
deep learning-recurrent structure
deep learning -any learning methor that can train a system with more than 2 or 3 non liear hidden layers
convolution neural network
sparse connectivity each neural only connects to part of the output of the previous layer
parameter sharing the neurons withdiffirent receptive fields can use the same set of parameters
less parameters than fully connected layer
deep learning
.deep learning
machine learning
why machine learning-recognizing patterns, recognizing anomalies, prediction
massive parallelism, huge data volumes stoae, data distribution, highspeed networks,etc
bigdata is every where.big data offers time advantage and data mining
mapreduce, hadoop, locality sensitive hashing
data parallel and model parallel
-volume- size and skill
velocity -geneerated at high speed
and variety, -variety like images, tables etc
veracity- uncertainty of the data
-it exceeds the processing capacity
big data is data that exceeds the processing capacity of conventional database systems. the data is too big , moves to fast or doesnt fit the structures of your databas
challenges in big data:
1) problem with storage
More data allows us
to see new aspects, see better aspects, see different aspects
more data..more information.
-examples facebook, an internet minute, smart phone users, big data is smilar to sall dat but bigger and requires diffirent approaches like techniques, tools and architctres to solve new problemms and old probles in abetter way.
it will bring challanges in volumes ,process and arigorithms
-what is the value of big data
-helps to track adays movement/estimate adays routine
-helps to know the emotion variations on national days
-helps with scene completion problems
data is mined so as to establish consumer emotion and trend.