[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]
2023-11: Warosu is now out of extended maintenance.

/sci/ - Science & Math


View post   

File: 85 KB, 714x582, 2FNLTensorFlow.png [View same] [iqdb] [saucenao] [google]
7706684 No.7706684 [Reply] [Original]

Who here is excited to work with TensorFlow after finals are finished? I'm new to making sense of big data and you may be too but Google has open sourced an easy to use suite of tools to help learn. Check it out if you're any type of scientist.

>> No.7706686
File: 374 KB, 252x448, tensors_flowing.gif [View same] [iqdb] [saucenao] [google]
7706686

>> No.7708318

I play on a vanilla WoW private serve that is emulated to original WoW with raids and eveything.

I'm wondering could I use tensorflow/neural networks to do a 40 man Molten Core/ BWL/ Naxx raid all done by a program learning it via wipes (trail and error)?

I would obviously have to teach DPS/Tanks/Healers how do do their roles

>> No.7709451

Bump

>> No.7709640

bump

>> No.7709643

>>7708318
'm relatively new as well and haven't played wow but lots of rs, and i'm going to guess probably not, unless you have excellent resources and time to train them

>> No.7709662

>>7709643

What about karazhan?

>> No.7709664

>>7706684
Tensor flow is just a library that does matrix operations and compute graphs in ways that are most relevant to machine learning fields. It gives you access to gpu acceleration and networked/distributed computation.

It is not necessarily the thing that is holding you back given that most RNN, LSTMs, ANNs have been coded up in torch, theano, and matlab already.

I am excited though.

>> No.7709915

I wish machine learning generals were a thing here, we could have a lot of fun

>> No.7709917 [DELETED] 

Good luck working, mate!

Read this and it will help you get through your days!

>>7709718

>> No.7709934

>>7706684
"Big Data" is such a meme phrase

>> No.7709940

>>7709934
>"Big Data" is such a meme phrase

90% of all data that exists today have been generated in the last two years.

Without big data all of it becomes useless.

>> No.7709955

>>7709940
So where's the line between normal data analysis and big data

>> No.7709962

>>7709955
When it's lots of data it's big data.

>> No.7709966

>>7709962
Why do we suddenly need a new word for it when data analysis suffices

>> No.7710699

>>7709966

Because data and big data is different.

Is salt water the only difference between a lake and an ocean?

>> No.7710711

>>7709962

If we can just have a bunch of people sit around calculating artillery tables why do we need electronic computers?