Forecasting European trade mark and design filings: An innovative approach including exogenous variables and IP offices' events

Use of the Industrial Property System in Colombia (2018): A Supervised Learning Application

“…In the case of supervised learning it was possible to identify 4 algorithms with a level of explanation higher than 80%, these are: (i) Linear Regression, with an elastic network regularization; (ii) Stochastic Gradient Descent, with Hinge loss function, Ringe regularization (L2) and a constant learning rate; (iii) Neural Networks, with 1,000 layers, with Adam’s solution algorithm and 2,000 iterations; (iv) Random Forest, with 10 trees. The results found in this study are consistent with those of [ 25 ], and some algorithms are added.…”

“…As for the applications of machine learning to intellectual property, in [ 24 ], they reviewed 57 papers on artificial intelligence, automatic and in-depth learning associated with intellectual property. In [ 25 ], the employed algorithms were Support Vector Machines, Neural Networks and Decision Trees.…”

Use of the Industrial Property System in Colombia (2018): A Supervised Learning Application

“…In the case of supervised learning it was possible to identify 4 algorithms with a level of explanation higher than 80%, these are: (i) Linear Regression, with an elastic network regularization; (ii) Stochastic Gradient Descent, with Hinge loss function, Ringe regularization (L2) and a constant learning rate; (iii) Neural Networks, with 1,000 layers, with Adam’s solution algorithm and 2,000 iterations; (iv) Random Forest, with 10 trees. The results found in this study are consistent with those of [ 25 ], and some algorithms are added.…”

“…As for the applications of machine learning to intellectual property, in [ 24 ], they reviewed 57 papers on artificial intelligence, automatic and in-depth learning associated with intellectual property. In [ 25 ], the employed algorithms were Support Vector Machines, Neural Networks and Decision Trees.…”

Effects of Clean Air Act on Patenting Activities in Chemical Industry: Learning from Past Experiences

“…The years where there is a lower numbers of grants are relatively many more when compared to the years with higher numbers of patents. Several studies have shown that it is possible to model the series of patents with ARIMA [41][42][43][44]. Some satisfactory model adjustments can be required to decrease the model residuals.…”

COVID-19 Pandemic and Technological Change: Analysis of Patent Applications

“…In the study, Dikta (2006) used time series analysis to forecast patent applications. Based on an investigation of multiple forecasting strategies employing European trademark and design registration data, emerging techniques based on artificial intelligence have outperformed traditional forecasting techniques (Havermans et al, 2017). Hingley and Park (2016) forecasted patent filings at the EPO using the Log-Linear Regression Model.…”

Use of the Industrial Property System in Colombia (2018): A Supervised Learning Application

“…In the case of supervised learning it was possible to identify 4 algorithms with a level of explanation higher than 80%, these are: (i) Linear Regression, with an elastic network regularization; (ii) Stochastic Gradient Descent, with Hinge loss function, Ringe regularization (L2) and a constant learning rate; (iii) Neural Networks, with 1,000 layers, with Adam’s solution algorithm and 2,000 iterations; (iv) Random Forest, with 10 trees. The results found in this study are consistent with those of [ 25 ], and some algorithms are added.…”

“…As for the applications of machine learning to intellectual property, in [ 24 ], they reviewed 57 papers on artificial intelligence, automatic and in-depth learning associated with intellectual property. In [ 25 ], the employed algorithms were Support Vector Machines, Neural Networks and Decision Trees.…”

Effects of Clean Air Act on Patenting Activities in Chemical Industry: Learning from Past Experiences

“…The years where there is a lower numbers of grants are relatively many more when compared to the years with higher numbers of patents. Several studies have shown that it is possible to model the series of patents with ARIMA [41][42][43][44]. Some satisfactory model adjustments can be required to decrease the model residuals.…”

COVID-19 Pandemic and Technological Change: Analysis of Patent Applications

“…In the study, Dikta (2006) used time series analysis to forecast patent applications. Based on an investigation of multiple forecasting strategies employing European trademark and design registration data, emerging techniques based on artificial intelligence have outperformed traditional forecasting techniques (Havermans et al, 2017). Hingley and Park (2016) forecasted patent filings at the EPO using the Log-Linear Regression Model.…”

Use of the Industrial Property System in Colombia (2018): A Supervised Learning Application

“…In the case of supervised learning it was possible to identify 4 algorithms with a level of explanation higher than 80%, these are: (i) Linear Regression, with an elastic network regularization; (ii) Stochastic Gradient Descent, with Hinge loss function, Ringe regularization (L2) and a constant learning rate; (iii) Neural Networks, with 1,000 layers, with Adam’s solution algorithm and 2,000 iterations; (iv) Random Forest, with 10 trees. The results found in this study are consistent with those of [ 25 ], and some algorithms are added.…”

“…As for the applications of machine learning to intellectual property, in [ 24 ], they reviewed 57 papers on artificial intelligence, automatic and in-depth learning associated with intellectual property. In [ 25 ], the employed algorithms were Support Vector Machines, Neural Networks and Decision Trees.…”

Effects of Clean Air Act on Patenting Activities in Chemical Industry: Learning from Past Experiences

“…The years where there is a lower numbers of grants are relatively many more when compared to the years with higher numbers of patents. Several studies have shown that it is possible to model the series of patents with ARIMA [41][42][43][44]. Some satisfactory model adjustments can be required to decrease the model residuals.…”

COVID-19 Pandemic and Technological Change: Analysis of Patent Applications

“…In the study, Dikta (2006) used time series analysis to forecast patent applications. Based on an investigation of multiple forecasting strategies employing European trademark and design registration data, emerging techniques based on artificial intelligence have outperformed traditional forecasting techniques (Havermans et al, 2017). Hingley and Park (2016) forecasted patent filings at the EPO using the Log-Linear Regression Model.…”