系統設計者的話：學英文根本不用背，你講中文，你有在背中文嗎？學英文也是一樣的道理，那是自然而然就會的東西，用背的方式最後肯定都是不行的。 更多...

最好的學習就是有英文的父母，不斷的跟你說話、耐心的教導你，時間長了，你也就慢慢就融入其中了。

但是我們沒有英文父母怎麼辦？看影片學英文就學的起來嗎？學不起來。

我們從小到大，也看過了不少英文電影，至少也看過1、200部也跑不掉，但是為什麼看了這麼多的英文電影，英文還是很爛呢？

主要的原因是我們沒有從基礎性的強化學習。舉個例子，西方的幼兒出生時，他至少聽了各種大人講英文至少聽了2年，才慢慢開始說英文、與大人互動。

而且都是一字一句慢慢的說、慢慢的教、漸進式的、慢慢的、越來越習慣、越來越聽的懂、越來越會說的過程。

那麼我們看的影片、電影過程中，都是大人成熟型的講話速度、方式、語調，我們怎麼可能跟的上？

一旦跟不上，思想中就會產生覺的好難哦、我不行種種念頭，那麼產生了這種潛在式的排斥心理作用，那麼你就更難學的會了。


所以我們必須打好基礎、深厚的基礎，萬丈高樓平地起，唯有打下深厚的基礎，後面你走的路才會越來越順。

程式界裏有句話，程式債欠的越多，最後就只能放棄。學英文也是一樣，英文單字債欠的越多，你就直接上來文法、句型、語調，最後也只能放棄。

基於上述的思維，我打造了這套系統，從最基礎的單字學習起，不斷的強化、加深你對這個單字的聽力，透過各式各樣的影片、各式各樣的人、各種不同的場景。

就只是為了訓練一個單字的聽力，無數的人跟你講同一個單字，講個20遍、50篇、100遍(且不重覆)，我就不信學不起來。


這套系統做出來後，我自己在使用的過程中，確實真的能聽出來他們講的這個單字，而且他們講的速度再快也能聽的出來。

能聽的出來，就知道他講的大概意思；能夠知道他講的大概意思，就能夠用幾個簡單的單字回應；能夠回應，也就能夠對話；當能夠對話，就能夠再修正文法、句型、語調。


我的目地就是希望自己至少能聽的懂人家在說什麼，用英文跟人家對話，大家共同努力、學習。若有其它想法或對這套系統設計的建議，也可以與我聯絡：ichich2013@gmail.com

Lian Li: Applying The Magic Of Neural Networks - JSConf.Asia 2016
觀看次數：409
about
neural
networks
today
how
they
github
on
google+
although
i'm
like
paul
gascoigne
and
he's
also
known
as
programmers
of
an
AI
in
like
an
rpg
game
counter-attack
or
maybe
you
want
to
action
for
example
its
deterministic
individual
players
does
it
doesn't
it
counter-attack
if
it's
between
25
and
seventy-five
percent
we
do
a
defense
for
decision-making
so
what
we
need
here
node
but
it's
not
called
neural
node
it's
called
neural
networks
so
we
have
to
like
understand
what
neural
networks
neuron
neuron
receives
input
through
its
dendrites
that's
a
fuzzy
purple
the
green
coordinate
and
inside
the
soma
gets
changed
in
a
deterministic
way
it
is
sent
through
the
axon
terminals
neuron
but
by
a
vast
network
of
neurons
deterministically
the
whole
thing
is
not
natural
neurons
with
artificial
neurons
magic
we
need
numerical
representations
that
or
algorithm
can
deduce
that
it's
implications
that
I
don't
want
to
for
potion
so
how
do
we
get
our
predictions
propagation
and
it's
called
that
because
start
with
our
inputs
that's
the
one
let's
focus
on
upper
node
so
it's
not
so
weights
randomly
and
now
we
have
to
like
nodes
to
the
input
nodes
and
let's
say
data
in
this
three-dimensional
get
there
mathematically
to
the
red
dot
to
know
how
neural
networks
work
but
chose
two
layers
with
six
nodes
because
it's
non
deterministic
and
those
algorithm
and
if
we
know
how
we're
doing
out
how
our
algorithm
performs
there's
we
predicted
it's
the
mean
squared
error
and
it's
out
how
our
algorithm
performs
with
data
case
we
just
use
the
cross-validation
algorithm
first
thing
is
adjusting
the
10000
iteration
in
the
100000th
cross-validation
error
seems
to
go
up
at
regularization
that
like
real
data
tributed
niels
bohr
but
i'm
not
quite
eventually
but
neural
networks
are
just
neural
networks
is
precisely
so
you
tensorflow
tender
flows
the
machine
i've
done
some
predictions
for
the

相關影片

加Line好友 每日8AM 推播1單字

加入好友

其他人正在學習的單字

1. deborah
2024-05-04 07:31
2. surround
2024-05-04 07:31
3. brain
2024-05-04 07:31
4. indebted
2024-05-04 07:31
5. payments
2024-05-04 07:31
6. which
2024-05-04 07:31
7. tangle
2024-05-04 07:31
8. joe
2024-05-04 07:31
9. food
2024-05-04 07:31
10. were
2024-05-04 07:31
11. h
2024-05-04 07:31
12. caves
2024-05-04 07:31
13. minutes
2024-05-04 07:31
14. some
2024-05-04 07:30
15. dinner
2024-05-04 07:30

本月觀看影片單字排行榜

本月查閱單字排行榜

本月聆聽單字排行榜