back when I was in the fifth grade my friends and I loved to play on the
friends and I loved to play on the
friends and I loved to play on the monkey bars we wouldn’t play on these
monkey bars we wouldn’t play on these
monkey bars we wouldn’t play on these monkey bars like any other kids but we
monkey bars like any other kids but we
monkey bars like any other kids but we try to do these crazy acrobatics on them
try to do these crazy acrobatics on them
try to do these crazy acrobatics on them jumping from the third bar to the
jumping from the third bar to the
jumping from the third bar to the seventh bar and trying to just do all
seventh bar and trying to just do all
seventh bar and trying to just do all these crazy flips and tricks but I
these crazy flips and tricks but I
these crazy flips and tricks but I remember one particular time this time I
remember one particular time this time I
remember one particular time this time I stepped up to the monkey bars got up and
stepped up to the monkey bars got up and
stepped up to the monkey bars got up and tried to swing for the 3rd monkey bar
tried to swing for the 3rd monkey bar
tried to swing for the 3rd monkey bar skipping for the first and second when I
skipping for the first and second when I
skipping for the first and second when I did this instead of latching on to the
did this instead of latching on to the
did this instead of latching on to the third one I ended up grabbing on to the
third one I ended up grabbing on to the
third one I ended up grabbing on to the air and seconds later I fell down to the
air and seconds later I fell down to the
air and seconds later I fell down to the ground turns out I had the wind knocked
ground turns out I had the wind knocked
ground turns out I had the wind knocked out of me and I was taken to the
out of me and I was taken to the
out of me and I was taken to the emergency room here I found out that I
emergency room here I found out that I
emergency room here I found out that I had 14 fractures in my back and that I’d
had 14 fractures in my back and that I’d
had 14 fractures in my back and that I’d have a month off from school now this is
have a month off from school now this is
have a month off from school now this is almost any fifth graders dream a month
almost any fifth graders dream a month
almost any fifth graders dream a month off and all I got to do was stay home
off and all I got to do was stay home
off and all I got to do was stay home and play a Nintendo DS but I also spent
and play a Nintendo DS but I also spent
and play a Nintendo DS but I also spent a lot of time thinking about my
a lot of time thinking about my
a lot of time thinking about my situation in my health I was grateful
situation in my health I was grateful
situation in my health I was grateful that I was going to heal but I realized
that I was going to heal but I realized
that I was going to heal but I realized that there was an alternative and it was
that there was an alternative and it was
that there was an alternative and it was really hard to ignore 5.4 million people
really hard to ignore 5.4 million people
really hard to ignore 5.4 million people in the United States are victims of some
in the United States are victims of some
in the United States are victims of some degree of paralysis and over 85 percent
degree of paralysis and over 85 percent
degree of paralysis and over 85 percent of them are unemployed meaning that
of them are unemployed meaning that
of them are unemployed meaning that they’re not able to have access to the
they’re not able to have access to the
they’re not able to have access to the greatest medical technologies that they
greatest medical technologies that they
greatest medical technologies that they need to stay relevant in terms of being
need to stay relevant in terms of being
need to stay relevant in terms of being able to carry out day-to-day tasks with
able to carry out day-to-day tasks with
able to carry out day-to-day tasks with this in mind with a couple of friends I
this in mind with a couple of friends I
this in mind with a couple of friends I spearheaded efforts to create a brain
spearheaded efforts to create a brain
spearheaded efforts to create a brain actuated exoskeleton now brain actuated
actuated exoskeleton now brain actuated
actuated exoskeleton now brain actuated exoskeleton that seems like a mouthful
exoskeleton that seems like a mouthful
exoskeleton that seems like a mouthful let’s break it down so the idea is that
let’s break it down so the idea is that
let’s break it down so the idea is that we have this robotic exoskeleton that
we have this robotic exoskeleton that
we have this robotic exoskeleton that goes on a physical therapy patient or a
goes on a physical therapy patient or a
goes on a physical therapy patient or a paralyzed patient that allows them to
paralyzed patient that allows them to
paralyzed patient that allows them to control their emotions based on just
control their emotions based on just
control their emotions based on just their electrical activity from their
their electrical activity from their
their electrical activity from their brain how you may be thinking what is
brain how you may be thinking what is
brain how you may be thinking what is this brain reading device it sounds like
this brain reading device it sounds like
this brain reading device it sounds like science fiction well the way we do this
science fiction well the way we do this
science fiction well the way we do this is through a powerful idea in computer
is through a powerful idea in computer
is through a powerful idea in computer science called machine learning machine
science called machine learning machine
science called machine learning machine learning is this idea that we can teach
learning is this idea that we can teach
learning is this idea that we can teach computers a specific generic task now I
computers a specific generic task now I
computers a specific generic task now I said specific and generic how does that
said specific and generic how does that
said specific and generic how does that work it’s specific in that the machine
work it’s specific in that the machine
work it’s specific in that the machine knows only how to do that one particular
knows only how to do that one particular
knows only how to do that one particular task but it knows how to handle a
task but it knows how to handle a
task but it knows how to handle a variety of different input cases and
variety of different input cases and
variety of different input cases and there are two main ways we can teach a
there are two main ways we can teach a
there are two main ways we can teach a computer this specific task the first is
computer this specific task the first is
computer this specific task the first is called supervised learning and it’s a
called supervised learning and it’s a
called supervised learning and it’s a lot like how humans read books when we
lot like how humans read books when we
lot like how humans read books when we read a book we try to understand the
read a book we try to understand the
read a book we try to understand the underlying motifs and download that data
underlying motifs and download that data
underlying motifs and download that data into our system and try to outline try
into our system and try to outline try
into our system and try to outline try to analyze the different patterns within
to analyze the different patterns within
to analyze the different patterns within it and when we do this machine’s on the
it and when we do this machine’s on the
it and when we do this machine’s on the other hand are able to take a look at
other hand are able to take a look at
other hand are able to take a look at large sets of data and understand how
large sets of data and understand how
large sets of data and understand how these underlying patterns exist in this
these underlying patterns exist in this
these underlying patterns exist in this data set the second is called
data set the second is called
data set the second is called unsupervised learning unsupervised
unsupervised learning unsupervised
unsupervised learning unsupervised learning is much like how humans explore
learning is much like how humans explore
learning is much like how humans explore like Christopher Columbus or Ponce de
like Christopher Columbus or Ponce de
like Christopher Columbus or Ponce de Leon the computer is placed in this
Leon the computer is placed in this
Leon the computer is placed in this virtual environment where you can give
virtual environment where you can give
virtual environment where you can give the computer certain rewards or negative
the computer certain rewards or negative
the computer certain rewards or negative reinforcements to understand how it will
reinforcements to understand how it will
reinforcements to understand how it will behave in this space the computer over
behave in this space the computer over
behave in this space the computer over time learns what is good and what is bad
time learns what is good and what is bad
time learns what is good and what is bad based on the reinforcement that it gets
based on the reinforcement that it gets
based on the reinforcement that it gets so with this in mind using the
so with this in mind using the
so with this in mind using the supervised learning approach our team
supervised learning approach our team
supervised learning approach our team built an EEG based machine learning
built an EEG based machine learning
built an EEG based machine learning algorithm now what this does is it takes
algorithm now what this does is it takes
algorithm now what this does is it takes the electrical potentials from the brain
the electrical potentials from the brain
the electrical potentials from the brain and understand understands the
and understand understands the
and understand understands the underlying patterns within them so that
underlying patterns within them so that
underlying patterns within them so that we’re able to execute different motions
we’re able to execute different motions
we’re able to execute different motions within the body now being able to do
within the body now being able to do
within the body now being able to do this allows for empowerment and access
this allows for empowerment and access
this allows for empowerment and access on a large scale it allows for access
on a large scale it allows for access
on a large scale it allows for access because we took a $50,000 exoskeleton
because we took a $50,000 exoskeleton
because we took a $50,000 exoskeleton and brought it down to just $600 and
and brought it down to just $600 and
and brought it down to just $600 and furthermore on top of this
furthermore on top of this
furthermore on top of this we’re bringing about another $50 version
we’re bringing about another $50 version
we’re bringing about another $50 version in the near future just based on this
in the near future just based on this
in the near future just based on this amazing machine learning based
amazing machine learning based
amazing machine learning based technology secondly machine learning is
technology secondly machine learning is
technology secondly machine learning is also making massive strides in terms of
also making massive strides in terms of
also making massive strides in terms of natural language processing natural
natural language processing natural
natural language processing natural language processing is this idea that we
language processing is this idea that we
language processing is this idea that we can start talking to computers and
can start talking to computers and
can start talking to computers and having natural interactions with them so
having natural interactions with them so
having natural interactions with them so this is much like Siri or Google now on
this is much like Siri or Google now on
this is much like Siri or Google now on your phones what this does is it allows
your phones what this does is it allows
your phones what this does is it allows us to carry out conversations and I
us to carry out conversations and I
us to carry out conversations and I think that this has massive implications
think that this has massive implications
think that this has massive implications in education a project that I started
in education a project that I started
in education a project that I started with a couple of friends is called a
with a couple of friends is called a
with a couple of friends is called a list education a list education allows
list education a list education allows
list education a list education allows for students to converse naturally with
for students to converse naturally with
for students to converse naturally with an online tutor what this allows
an online tutor what this allows
an online tutor what this allows students to do is it actually allows
students to do is it actually allows
students to do is it actually allows them to ask questions much that much
them to ask questions much that much
them to ask questions much that much like of those that they would ask a real
like of those that they would ask a real
like of those that they would ask a real tutor for example I could ask the
tutor for example I could ask the
tutor for example I could ask the textbook what is the complex number I
textbook what is the complex number I
textbook what is the complex number I and the textbook would be able to
and the textbook would be able to
and the textbook would be able to respond to me that I is the square root
respond to me that I is the square root
respond to me that I is the square root of negative 1 but on top of just
of negative 1 but on top of just
of negative 1 but on top of just answering these fact-based questions the
answering these fact-based questions the
answering these fact-based questions the system is also able to understand how
system is also able to understand how
system is also able to understand how much the student has learned so at one
much the student has learned so at one
much the student has learned so at one point when the student thinks that they
point when the student thinks that they
point when the student thinks that they are done this system will go back and
are done this system will go back and
are done this system will go back and ask them questions when the system asks
ask them questions when the system asks
ask them questions when the system asks these questions the system is actually
these questions the system is actually
these questions the system is actually assessing how much the student has
assessing how much the student has
assessing how much the student has learned throughout the process and if
learned throughout the process and if
learned throughout the process and if they haven’t understood the concept
they haven’t understood the concept
they haven’t understood the concept fully this the system provides the
fully this the system provides the
fully this the system provides the student supplementary material in the
student supplementary material in the
student supplementary material in the form of voice much like a real tutor
form of voice much like a real tutor
form of voice much like a real tutor does and what this is allowing for is
does and what this is allowing for is
does and what this is allowing for is empowerment we’re empowering people all
empowerment we’re empowering people all
empowerment we’re empowering people all over the world to have access to a
over the world to have access to a
over the world to have access to a quality education no matter their
quality education no matter their
quality education no matter their background and no matter where they live
background and no matter where they live
background and no matter where they live lastly I think machine learning has
lastly I think machine learning has
lastly I think machine learning has massive implications also on a more
massive implications also on a more
massive implications also on a more global level here we see a map of a
global level here we see a map of a
global level here we see a map of a hurricane forecast system and you can
hurricane forecast system and you can
hurricane forecast system and you can see it’s a complete mess this past
see it’s a complete mess this past
see it’s a complete mess this past hurricane season for us was one of the
hurricane season for us was one of the
hurricane season for us was one of the most was one of the worst that we’ve
most was one of the worst that we’ve
most was one of the worst that we’ve seen in a long time and
seen in a long time and
seen in a long time and wanting to understand these sort of
wanting to understand these sort of
wanting to understand these sort of interactions of how climate science
interactions of how climate science
interactions of how climate science works I tried to apply a machine
works I tried to apply a machine
works I tried to apply a machine learning approach to hurricanes and the
learning approach to hurricanes and the
learning approach to hurricanes and the way I did this is by creating a system
way I did this is by creating a system
way I did this is by creating a system that allows for intensity and forecast
that allows for intensity and forecast
that allows for intensity and forecast analysis so that we know where the
analysis so that we know where the
analysis so that we know where the hurricane is at any given time and what
hurricane is at any given time and what
hurricane is at any given time and what its particular intensity would be by
its particular intensity would be by
its particular intensity would be by knowing this information we can inform
knowing this information we can inform
knowing this information we can inform we can inform people that need to make
we can inform people that need to make
we can inform people that need to make their decisions of the various different
their decisions of the various different
their decisions of the various different types of implications of the hurricane
types of implications of the hurricane
types of implications of the hurricane that’s impending on top of this it’s
that’s impending on top of this it’s
that’s impending on top of this it’s important to note that when I started
important to note that when I started
important to note that when I started this project I had zero climate science
this project I had zero climate science
this project I had zero climate science experience other than some quick
experience other than some quick
experience other than some quick research that I did on how these systems
research that I did on how these systems
research that I did on how these systems work all I all I knew was that there was
work all I all I knew was that there was
work all I all I knew was that there was this data set and that I needed to find
this data set and that I needed to find
this data set and that I needed to find out where the hurricane was going to be
out where the hurricane was going to be
out where the hurricane was going to be at any given time using this supervised
at any given time using this supervised
at any given time using this supervised learning approach I understood that
learning approach I understood that
learning approach I understood that machine learning helped me be timely in
machine learning helped me be timely in
machine learning helped me be timely in that it allowed me to create a system on
that it allowed me to create a system on
that it allowed me to create a system on a subject that I had no experience about
a subject that I had no experience about
a subject that I had no experience about in the past
in the past
in the past this shows that machine learning can
this shows that machine learning can
this shows that machine learning can also serve to inspire others through
also serve to inspire others through
also serve to inspire others through various projects and various endeavors
various projects and various endeavors
various projects and various endeavors we can inspire the nation’s youth to
we can inspire the nation’s youth to
we can inspire the nation’s youth to pursue computer science and to pursue
pursue computer science and to pursue
pursue computer science and to pursue these activities but on top of that I
these activities but on top of that I
these activities but on top of that I ask the audience to contribute to data
ask the audience to contribute to data
ask the audience to contribute to data sets because every time that there is
sets because every time that there is
sets because every time that there is new data available we’re increasing our
new data available we’re increasing our
new data available we’re increasing our generalization of these machine learning
generalization of these machine learning
generalization of these machine learning systems furthermore I’d like to end with
systems furthermore I’d like to end with
systems furthermore I’d like to end with the fact that machine learning is
the fact that machine learning is
the fact that machine learning is allowing for us to have an amazing and
allowing for us to have an amazing and
allowing for us to have an amazing and empowering future and it’s allowing us
empowering future and it’s allowing us
empowering future and it’s allowing us to talk about projects and ideas that
to talk about projects and ideas that
to talk about projects and ideas that help change the world thank you
help change the world thank you
help change the world thank you you
”
Be First to Comment