well up thorax unis fasciitis is a
species of ants that can sort through
many objects objects such as eggs larvae
dance or anything else in at may
encounter essentially the ants use a
decentralized system every single ant
will make a decision based off its local
area for example a gnat may approach an
area and see that there’s lots of larvae
but only one egg so it’ll pick up this
egg and drop it off somewhere else where
it sees a lot more eggs this process
occurs over and over and over again
forming an overarching layer of
intelligence what you’ll notice is that
clusters of similar items will form near
each other for example a cluster of
smaller X will be beside a cluster of
larger eggs and then a bit off in the
distance you’ll see clusters of larvae
arranged in their own subsets the ass
has a very dynamic clustering system and
what’s really intriguing about it is how
the overarching layer of intelligence is
built by effectively using the local
intelligence for every single ant
this has really intrigued scientists and
researchers as well in fact it’s even
prompted them to come up with an
algorithm which mimics this behavior
this algorithm is called ant brood
sorting but can you imagine it
optimizing your financial portfolio well
that’s exactly what I did last year
Modern Portfolio theory is the concept
that most fund managers will use to
optimize their financial portfolios
essentially it states that an optimal
portfolio will do two things it’ll give
you as much returns as possible but
it’ll also take on the least amount of
risk that
all that really changes among fund
managers is the models methods and
formulas that we’ll use to incorporate
this very basic principle into their
financial portfolios I wanted to create
my own formula and I wanted it to be
fully automated so that I could remove
all of the human bias and to do that I
looked at existing computational
algorithms neural networks are an
algorithm that’s based off your brain
they’re basically a set of artificial
neurons they’ll communicate with each
other to take a set of inputs and give
you a desired output and just like the
brain they know what to do by learning
off of older examples I wanted to use
neural networks to predict the future
price of a stock which gives me future
returns but what about risk risk is an
extremely important part about
optimizing a financial portfolio and how
do you reduce it well there are many
ways but it all comes down to the old
adage don’t put all your eggs in one
basket to not put all my eggs in one
basket
I needed to identify which stocks are
inherently similar so that I could avoid
picking them and to do this I used the
ant brood of sorting algorithm and my
overarching algorithm worked after
testing in the years of 2013 the 2016 I
learned that in the time that market
went a 50% my algorithm went up a full
87% and by looking at beta levels which
are a financial tool to help assess risk
levels I learned that my algorithm
generated portfolios with a much lower
risk level so the algorithm worked and
that means that the two individual
concepts in it must have as well and
there’s one thing that connects these
two individual concepts they’re both
based off of nature
nature is something that’s far from
perfect but it was still able to inspire
scientists and researchers into
developing these algorithms and it’s not
just neural networks and attribute
sorting it’s also artificial be calling
the algorithm Firefly algorithms
gravitational search algorithms genetic
algorithms and so many more this is
formed a field known as nature-inspired
computing nature-inspired computing
refers to any computational algorithm
that has been inspired by nature or
system inside of nature you see in
nature there are many complex problems
and nature often finds a way to solve
them for example your brain is talking
on issues on a daily basis nonetheless
or evolution has solved many problems
phenomena like these have really
captivated the attention of scientists
and researchers and so they’ve created
algorithms which mimic them but why why
are scientists and researchers devoting
so much time and energy into creating
these algorithms what makes them so
special well the answer to that question
lies in the simple fact that nature is
unpredictable as a result whenever an
animal or species in nature has been
presented with a problem it’s had to
adapt Nature has had to use its
resources in remarkable ways in order to
solve these problems
an example is the artificial bee colony
algorithm this algorithm is based off
the foraging method that bees will use
in nature despite the fact that bees
aren’t exactly the most intelligent
creatures out there they’re still able
to create a very effective foraging
method essentially what they’ve done is
they formed a hive light the bees have
combined their intelligence and their
resources into coming up with an
adaptable foraging method that works in
a variety of location
and the ability to adapt is the basis of
future algorithms traditional rule based
computing can’t really solve many
problems effectively without requiring
lots of human intervention because for
the majority of cases problems don’t
really obliged to one standard rule so
if nature inspired computing can adapt
the nature and smart computing can solve
many problems and it’s done just that
the after mention artificial bee colony
algorithm has been used in a variety of
locations one notable example is the
detection of breast cancer samples
researchers from the islamic azad
university in iran attempted to use the
artificial bee colony algorithm in order
to detect breast cancer samples and they
did so at a success rate of 96.5% that
is incredibly competitive through
existing methods but it’s also a much
simpler model bees are detecting breast
cancer bees are saving lives and for
another example we can look at the
particle swarm optimization algorithm
which has been used by researchers at
the Electric Power College in South
China’s University of Technology the
particle swarm optimization algorithm
essentially refers to a nature-inspired
algorithm that has been designed to
simulate social behavior by using this
algorithm to solve the heat system
planning problem which is basically
referring to trying to optimize your
heating system so it’s a lot more
eco-friendly and a lot more economically
efficient they were able to create a
heating system that when compared to
traditional systems was a lot more
eco-friendly and a lot more economically
efficient but yet another area where
nature inspired computing has worked is
sports unanimous AI is a company that’s
developed a technology called soir
a I swarm AI is a set of interfaces and
algorithms designed to simulate swarms
in nature they believe that if they can
combine humans in the same way that a
swarm of bees fish birds ants or any
other swarm in nature can they’ll be
able to accomplish remarkable things and
it’s really hard to argue that they
failed for example not only did they
predict that the New England Patriots
would defeat the Atlanta Falcons in
Super Bowl 51 but they also predicted
the exact score of 34 to 28 before the
game plus they predicted the top four
horses in the Kentucky Derby and in
order so they took a $20 bet and turned
it into a 11 thousand eight hundred
dollar payout plus they’ve predicted the
Chicago Cubs to defeat the Cleveland
Indians and the World Series of fear is
back and for the Pittsburgh Penguins to
win the Stanley Cup back in 2016 but it
isn’t the fact that they’ve been able to
accurately predict these events that’s
so very impressive with random events
like these obviously they won’t be a
hundred percent accurate all of the time
rather what’s really intriguing is how
they’re able to take a set of average
fans and make them I’ll perform the
experts and they’re doing it in the same
way that nature does so what do these
examples prove well these examples show
us that nature is moving forward the
days of traditional rule based computing
are coming to an end now we’re using
innovative solutions to solve problems
problems like detecting breast cancer
optimizing your heating system or
predicting the next Super Bowl winner
all at you sir acquired humans there was
simply no other conceivable way to do
these things luckily people challenge
themselves they challenged themselves to
think
innovative solutions solutions that
would have been dismissed as trivial or
unpractical before these people showed
us that there aren’t boundaries when it
comes to innovative thinking biologists
and computer scientists may seem like
polar opposites one study something very
natural the other something man-made
however that didn’t stop these two
fields from joining together people took
a step outside of their area of
expertise to solve problems that were in
their area of expertise and to move
forward will have to join together our
areas of knowledge and our skills and
that’s precisely what I challenge you
all to do thank you [Applause]
species of ants that can sort through
many objects objects such as eggs larvae
dance or anything else in at may
encounter essentially the ants use a
decentralized system every single ant
will make a decision based off its local
area for example a gnat may approach an
area and see that there’s lots of larvae
but only one egg so it’ll pick up this
egg and drop it off somewhere else where
it sees a lot more eggs this process
occurs over and over and over again
forming an overarching layer of
intelligence what you’ll notice is that
clusters of similar items will form near
each other for example a cluster of
smaller X will be beside a cluster of
larger eggs and then a bit off in the
distance you’ll see clusters of larvae
arranged in their own subsets the ass
has a very dynamic clustering system and
what’s really intriguing about it is how
the overarching layer of intelligence is
built by effectively using the local
intelligence for every single ant
this has really intrigued scientists and
researchers as well in fact it’s even
prompted them to come up with an
algorithm which mimics this behavior
this algorithm is called ant brood
sorting but can you imagine it
optimizing your financial portfolio well
that’s exactly what I did last year
Modern Portfolio theory is the concept
that most fund managers will use to
optimize their financial portfolios
essentially it states that an optimal
portfolio will do two things it’ll give
you as much returns as possible but
it’ll also take on the least amount of
risk that
all that really changes among fund
managers is the models methods and
formulas that we’ll use to incorporate
this very basic principle into their
financial portfolios I wanted to create
my own formula and I wanted it to be
fully automated so that I could remove
all of the human bias and to do that I
looked at existing computational
algorithms neural networks are an
algorithm that’s based off your brain
they’re basically a set of artificial
neurons they’ll communicate with each
other to take a set of inputs and give
you a desired output and just like the
brain they know what to do by learning
off of older examples I wanted to use
neural networks to predict the future
price of a stock which gives me future
returns but what about risk risk is an
extremely important part about
optimizing a financial portfolio and how
do you reduce it well there are many
ways but it all comes down to the old
adage don’t put all your eggs in one
basket to not put all my eggs in one
basket
I needed to identify which stocks are
inherently similar so that I could avoid
picking them and to do this I used the
ant brood of sorting algorithm and my
overarching algorithm worked after
testing in the years of 2013 the 2016 I
learned that in the time that market
went a 50% my algorithm went up a full
87% and by looking at beta levels which
are a financial tool to help assess risk
levels I learned that my algorithm
generated portfolios with a much lower
risk level so the algorithm worked and
that means that the two individual
concepts in it must have as well and
there’s one thing that connects these
two individual concepts they’re both
based off of nature
nature is something that’s far from
perfect but it was still able to inspire
scientists and researchers into
developing these algorithms and it’s not
just neural networks and attribute
sorting it’s also artificial be calling
the algorithm Firefly algorithms
gravitational search algorithms genetic
algorithms and so many more this is
formed a field known as nature-inspired
computing nature-inspired computing
refers to any computational algorithm
that has been inspired by nature or
system inside of nature you see in
nature there are many complex problems
and nature often finds a way to solve
them for example your brain is talking
on issues on a daily basis nonetheless
or evolution has solved many problems
phenomena like these have really
captivated the attention of scientists
and researchers and so they’ve created
algorithms which mimic them but why why
are scientists and researchers devoting
so much time and energy into creating
these algorithms what makes them so
special well the answer to that question
lies in the simple fact that nature is
unpredictable as a result whenever an
animal or species in nature has been
presented with a problem it’s had to
adapt Nature has had to use its
resources in remarkable ways in order to
solve these problems
an example is the artificial bee colony
algorithm this algorithm is based off
the foraging method that bees will use
in nature despite the fact that bees
aren’t exactly the most intelligent
creatures out there they’re still able
to create a very effective foraging
method essentially what they’ve done is
they formed a hive light the bees have
combined their intelligence and their
resources into coming up with an
adaptable foraging method that works in
a variety of location
and the ability to adapt is the basis of
future algorithms traditional rule based
computing can’t really solve many
problems effectively without requiring
lots of human intervention because for
the majority of cases problems don’t
really obliged to one standard rule so
if nature inspired computing can adapt
the nature and smart computing can solve
many problems and it’s done just that
the after mention artificial bee colony
algorithm has been used in a variety of
locations one notable example is the
detection of breast cancer samples
researchers from the islamic azad
university in iran attempted to use the
artificial bee colony algorithm in order
to detect breast cancer samples and they
did so at a success rate of 96.5% that
is incredibly competitive through
existing methods but it’s also a much
simpler model bees are detecting breast
cancer bees are saving lives and for
another example we can look at the
particle swarm optimization algorithm
which has been used by researchers at
the Electric Power College in South
China’s University of Technology the
particle swarm optimization algorithm
essentially refers to a nature-inspired
algorithm that has been designed to
simulate social behavior by using this
algorithm to solve the heat system
planning problem which is basically
referring to trying to optimize your
heating system so it’s a lot more
eco-friendly and a lot more economically
efficient they were able to create a
heating system that when compared to
traditional systems was a lot more
eco-friendly and a lot more economically
efficient but yet another area where
nature inspired computing has worked is
sports unanimous AI is a company that’s
developed a technology called soir
a I swarm AI is a set of interfaces and
algorithms designed to simulate swarms
in nature they believe that if they can
combine humans in the same way that a
swarm of bees fish birds ants or any
other swarm in nature can they’ll be
able to accomplish remarkable things and
it’s really hard to argue that they
failed for example not only did they
predict that the New England Patriots
would defeat the Atlanta Falcons in
Super Bowl 51 but they also predicted
the exact score of 34 to 28 before the
game plus they predicted the top four
horses in the Kentucky Derby and in
order so they took a $20 bet and turned
it into a 11 thousand eight hundred
dollar payout plus they’ve predicted the
Chicago Cubs to defeat the Cleveland
Indians and the World Series of fear is
back and for the Pittsburgh Penguins to
win the Stanley Cup back in 2016 but it
isn’t the fact that they’ve been able to
accurately predict these events that’s
so very impressive with random events
like these obviously they won’t be a
hundred percent accurate all of the time
rather what’s really intriguing is how
they’re able to take a set of average
fans and make them I’ll perform the
experts and they’re doing it in the same
way that nature does so what do these
examples prove well these examples show
us that nature is moving forward the
days of traditional rule based computing
are coming to an end now we’re using
innovative solutions to solve problems
problems like detecting breast cancer
optimizing your heating system or
predicting the next Super Bowl winner
all at you sir acquired humans there was
simply no other conceivable way to do
these things luckily people challenge
themselves they challenged themselves to
think
innovative solutions solutions that
would have been dismissed as trivial or
unpractical before these people showed
us that there aren’t boundaries when it
comes to innovative thinking biologists
and computer scientists may seem like
polar opposites one study something very
natural the other something man-made
however that didn’t stop these two
fields from joining together people took
a step outside of their area of
expertise to solve problems that were in
their area of expertise and to move
forward will have to join together our
areas of knowledge and our skills and
that’s precisely what I challenge you
all to do thank you [Applause]
Please follow and like us: