machine learning - Neural network feature combinatorics -


suppose have neural network sufficiently high number of hidden layers, hidden units per layer, , iterations, such these parameters not affect network's predictions.

given features x1, x2, ..., xn, possible (to prove) whether or not range of potential features redundant given subset of features (x1 through xn). is, given these features (x1 through xn), can neural network discern other features such as:

  • differences or additions (x1-x49, or x17+xn)?
  • products , ratios (x1*x1, or x47/xn)
  • higher order polynomials (or products of sequences ∏(x1 through xn))
  • trigonometric functions based upon original features (sin(x1*xn) + x17)
  • logarithmic functions (ln(x2*x4)/x6)

it in line of inquiry wondering if there situations using neural network need add higher order, or different, functions network predict accurately.

in general, given adequate number of features, possible network model any graph, , if not functional domains can neural networks not predict?

furthermore, there research point out references topic?

thanks!

given features x1, x2, ..., xn, possible (to prove) whether or not range of potential features redundant given subset of features (x1 through xn). is, given these features (x1 through xn), can neural network discern other features

it seems if looking dimensionality reduction neural networks. autoencoders can that:

  • you have inputs x1, x2, ..., xn.
  • you create network gets inputs (n input nodes). has hidden layers, bottleneck (k nodes, k < n) , output layer (n nodes).
  • the target recreate input.

when trained, discard layers after output. if network able restore inputs bottleneck, later layers not necessary.

in general, given adequate number of features, possible network model graph, , if not functional domains can neural networks not predict?

i guess looking universal approximation theorem. in short: neural networks can approximate continuous functions on compact subsets of r^n arbitrary close long give them enough nodes , @ least 1 hidden layer


Comments

Post a Comment

Popular posts from this blog

android - InAppBilling registering BroadcastReceiver in AndroidManifest -

i'm implementing inapp billing in android app. works well, however, i'm trying decouple broadcast receiver activity manifest. suggestion in android's trivialdrive sample: // important: dynamically register broadcast messages updated purchases. // register receiver here instead of <receiver> in manifest // because call getpurchases() @ startup, therefore can ignore // broadcasts sent while app isn't running. // note: registering listener in activity bad idea, done here // because sample. regardless, receiver must registered after // iabhelper setup, before first call getpurchases(). currently there's class extends broadcastreceiver : public class iabbroadcastreceiver extends broadcastreceiver { /** * intent action receiver should filter for. */ public static final string action = "com.android.vending.billing.purchases_updated"; private final iabbroadcastlistener mlistener; public iabbroadcastreceiver(iabbroadcastlistener listener) { m...
Read more

python Tkinter Capturing keyboard events save as one single string -

is possible store keyboard events 1 string? code below stores 1 char , prints it. card reader or bar code reader, contains collection of character/ string not 1 character @ time. goal save char pressed text variable. from tkinter import * root = tk() def key(event): text= event.char text+= event.char print ("pressed", text) def callback(event): frame.focus_set() print ("clicked at", event.x, event.y) frame = frame(root, width=100, height=100) frame.bind("<key>", key) frame.bind("<button-1>", callback) frame.pack() root.mainloop() currently, creating text variable , everytime key() function gets called, , text stores last character typed. you can define text module level variable , , use module level text inside key function - text = '' def key(event): global text text+= event.char print("pressed", text)
Read more

sql server - Why does Linq-to-SQL add unnecessary COUNT()? -

here tables: create table customer ( cid int primary key, name varchar(32) ) create table ord ( oid int primary key, cid int foreign key references customer, address varchar(20) ) here's linq-to-sql statement: var aaa = c in db.customer select new { c, o = c.ord.tolist() }; and here's inexplicable query generated linq-to-sql: select [t0].[cid] [cid], [t0].[name] [name], [t1].[oid] [oid], [t1].[cid] [cid2], [t1].[address] [address], (select count(*) [dbo].[ord] [t2] [t2].[cid] = [t0].[cid]) [value] [dbo].[customer] [t0] left outer join [dbo].[ord] [t1] on [t1].[cid] = [t0].[cid] order [t0].[cid], [t1].[oid] i want understand how rid of count(*) part. it's entirely uncalled for! linq needs count(*) in order determine number of entries returned o = c.ord.tolist() .
Read more